def png2coco(INFO,LICENSES,CATEGORIES,output_filepath,dataset):
coco_output = {
"info":INFO,
"licenses":LICENSES,
"categories":CATEGORIES,
"images":[],
"annotations":[]
}
image_id = 1
segmentation_id = 1
for i in range(len(dataset)):
print(i,'/',len(dataset))
image_files = listdir(dataset[i])
for j in range(len(image_files)):
if 'image' in image_files[j]:
image_filename = join(dataset[i],image_files[j])
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
i,image_filename,image.size)
coco_output["images"].append(image_info)
else:
annotation_filename = join(dataset[i],image_files[j])
class_id = [x['id'] for x in CATEGORIES if x['name'] in image_files[j]][0]
category_info = {'id':class_id,'is_crowd':'crowd' in image_filename}
binary_mask = np.asarray(Image.open(annotation_filename)
.convert('1')).astype(np.uint8)
annotation_info = pycococreatortools.create_annotation_info(
i,image_id,category_info,binary_mask,
image.size,tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
with open(output_filepath,'w') as output_json_file:
json.dump(coco_output,output_json_file)
def main():
coco_output = {"images": [], "annotations": []}
image_id = 1
segmentation_id = 1
# filter for jpeg images
for root, _, files in os.walk(IMAGE_DIR):
image_files = filter_for_jpeg(root, files) # all image name
# go through each image
for image_filename in image_files:
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size)
coco_output["images"].append(image_info)
# filter for associated png annotations
for root, _, files in os.walk(ANNOTATION_DIR):
annotation_files = filter_for_annotations(
root, files, image_filename) # all this image files
# go through each associated annotation
for annotation_filename in annotation_files:
print(annotation_filename)
class_id = [
x['id'] for x in CATEGORIES
if x['name'] in annotation_filename
][0]
category_info = {
'id': class_id,
'is_crowd': 'crowd' in image_filename
}
binary_mask = np.asarray(
Image.open(annotation_filename).convert('1')).astype(
np.uint8)
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id,
image_id,
category_info,
binary_mask,
image.size,
tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
with open(
ROOT_DIR + '/amodel/frame{}.json'.format(
os.path.basename(image_filename)[:-4]),
'w') as output_json_file:
json.dump(coco_output, output_json_file)
def main():
# json dictionary
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_id = 1
segmentation_id = 1
for root, _, files in os.walk(IMAGE_DIR):
image_paths = filter_for_jpeg(root, files)
num_of_image_files = len(image_paths)
for image_path in image_paths:
image = Image.open(image_path)
image_name = os.path.basename(image_path)
image_info = pycococreatortools.create_image_info(
image_id, image_name, image.size)
coco_output["images"].append(image_info)
# get maskes
rle_masks = df.loc[df['ImageId'] == image_name,
'EncodedPixels'].tolist()
num_of_rle_masks = len(rle_masks)
for index in range(num_of_rle_masks):
binary_mask = rle_decode(rle_masks[index])
class_id = 1 # all images are 1, with ship
category_info = {'id': class_id, 'is_crowd': 0}
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id,
image_id,
category_info,
binary_mask,
image.size,
tolerance=2)
# filter out low-quality annotations
# save low-quality annotations for later
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
else:
save_bad_ann(image_name, binary_mask, segmentation_id)
# assign an id to each annotation (no matter if it's filtered out)
segmentation_id = segmentation_id + 1
if not image_id % 100:
print("%d of %d is done." % (image_id, num_of_image_files))
image_id = image_id + 1
with open('../data/annotations/train_ship_instances_v2.json',
'w') as output_json_file:
# json.dump(coco_output, output_json_file)
json.dump(coco_output, output_json_file, indent=4)
def coco_img_info(self):
img_info = pycococreatortools.create_image_info(
self.img_id,
self.img_path,
self.img_size
)
return img_info
def main():
# 最终放进json文件里的字典
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [], # 放一个空列表占位置，后面再append
"annotations": []
}
image_id = 1
segmentation_id = 1
# 最外层的循环是图片，因为图片的基本信息需要共享
# IMAGE_DIR路径下找到所有的图片
for root, _, files in os.walk(IMAGE_DIR):
image_paths = filter_for_jpeg(root, files) # 图片文件地址
num_of_image_files = len(image_paths) # 图片个数
# 遍历每一张图片
for image_path in image_paths:
# 提取图片信息
image = Image.open(image_path)
image_name = os.path.basename(image_path) # 不需要具体的路径，只要图片文件名
image_info = pycococreatortools.create_image_info(
image_id, image_name, image.size)
coco_output["images"].append(image_info)
# 内层循环是mask，把每一张图片的mask搜索出来
rle_masks = df.loc[df['ImageId'] == image_name, 'EncodedPixels'].tolist()
num_of_rle_masks = len(rle_masks)
for index in range(num_of_rle_masks):
binary_mask = rle_decode(rle_masks[index])
class_id = 1 # 所有图片的类别都是1，ship
category_info = {'id': class_id, 'is_crowd': 0}
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id, image_id, category_info, binary_mask,
image.size, tolerance=2)
# 不是所有的标注都会被转换,低质量标注会被过滤掉
# 正常的标注加入数据集，不好的标注保存供观察
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
else:
save_bad_ann(image_name, binary_mask, segmentation_id)
# 无论标注是否被写入数据集，均分配一个编号
segmentation_id = segmentation_id + 1
print("%d of %d is done."%(image_id,num_of_image_files))
image_id = image_id + 1
with open('{}/annotations/instances_ships_train2018.json'.format(ROOT_DIR), 'w') as output_json_file:
# json.dump(coco_output, output_json_file)
json.dump(coco_output, output_json_file,indent=4)
def create_annotations():
print("Started processing.")
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_id = 1
segmentation_id = 1
for root, _, files in os.walk(IMAGE_DIR):
image_paths = filter_for_jpeg(root, files)
num_of_image_files = len(image_paths)
for image_path in image_paths:
image = Image.open(image_path)
image_name = os.path.basename(image_path)
image_info = pycococreatortools.create_image_info(
image_id, image_name, image.size)
coco_output["images"].append(image_info)
rle_masks = df.loc[df['ImageId'] == image_name,
'EncodedPixels'].tolist()
num_of_rle_masks = len(rle_masks)
for index in range(num_of_rle_masks):
binary_mask = rle_decode(rle_masks[index])
class_id = 1
category_info = {'id': class_id, 'is_crowd': 0}
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id,
image_id,
category_info,
binary_mask,
image.size,
tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
else:
save_bad_ann(image_name, binary_mask, segmentation_id)
segmentation_id = segmentation_id + 1
if (image_id % 1000 == 0):
print("Processing %d of %d is done. %d perc." %
(image_id, num_of_image_files,
np.round(image_id / num_of_image_files, 2) * 100))
image_id = image_id + 1
return coco_output
def gen_coco():
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_id = 1
segmentation_id = 1
# filter for jpeg images
for root, _, files in os.walk(IMAGE_DIR):
image_files = filter_for_image(root, files)
# 如果是肿瘤图片，直接添加。
# 如果是正常图片，先收拢一起，然后随机采样和肿瘤图片一样数量的，保正负样本均衡。
tumor_image_files = [image_filename for image_filename in image_files if 'tumor' in os.path.basename(image_filename)]
normal_image_files = random.sample([image_filename for image_filename in image_files if 'normal' in os.path.basename(image_filename)],len(tumor_image_files))
# go through each image
for image_filename in tqdm(tumor_image_files+normal_image_files):
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size)
coco_output["images"].append(image_info)
# filter for associated png annotations
for root, _, files in os.walk(MASK_DIR):
annotation_files = filter_for_annotations(root, files, image_filename)
# go through each associated annotation
for annotation_filename in annotation_files:
class_id = [x['id'] for x in CATEGORIES if x['name'] in os.path.basename(annotation_filename)][0]
binary_mask = np.asarray(Image.open(annotation_filename).convert('1')).astype(np.uint8)
# class_id = 1 if binary_mask.sum() > 0 else 0
category_info = {'id': class_id, 'is_crowd': 0}
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id, image_id, category_info, binary_mask,
image.size, tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
with open('{}/instances_camelyon16_train2020.json'.format(ROOT_DIR), 'w') as output_json_file:
json.dump(coco_output, output_json_file)
def run(args):
infer_dataset = dataloader.VOC12ImageDataset(args.infer_list,
voc12_root=args.voc12_root)
infer_data_loader = DataLoader(infer_dataset,
shuffle=False,
num_workers=args.num_workers,
pin_memory=True)
val_json = json.load(
open(os.path.join(VOC2012_JSON_FOLDER, 'pascal_val2012.json')))
# Do not use this file for evaluation!
coco_output = {}
coco_output["images"] = []
coco_output["annotations"] = []
coco_output['categories'] = val_json['categories']
coco_output['type'] = val_json['type']
for iter, pack in tqdm(enumerate(infer_data_loader),
total=len(infer_dataset)):
img_name = pack['name'][0]
img_id = int(img_name[:4] + img_name[5:])
img_size = pack['img'].shape[2:]
image_info = pycococreatortools.create_image_info(
img_id, img_name + ".jpg", (img_size[1], img_size[0]))
coco_output["images"].append(image_info)
ann = np.load(os.path.join(args.ins_seg_out_dir, img_name) + '.npy',
allow_pickle=True).item()
instance_id = 1
for score, mask, class_id in zip(ann['score'], ann['mask'],
ann['class']):
if score < 1e-5:
continue
category_info = {'id': class_id, 'is_crowd': False}
annotation_info = pycococreatortools.create_annotation_info(
instance_id,
img_id,
category_info,
mask,
img_size[::-1],
tolerance=0)
instance_id += 1
coco_output['annotations'].append(annotation_info)
with open('voc2012_train_custom.json', 'w') as outfile:
json.dump(coco_output, outfile)
def main():
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_id = 0
# segmentation_id = 1
# filter for jpeg images
for root, _, files in os.walk(IMAGE_DIR):
image_files = filter_for_jpeg(root, files)
for root, _, files in os.walk(mask_DIR):
image_GT = filter_for_jpeg(root, files)
# for image_filename in image_files:
for num in range(len(image_files)):
image = cv2.imread(image_files[num])
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_files[num]), image.shape)
print(image.shape)
coco_output["images"].append(image_info)
# image_name = image_files[num].split("/")[-1]
f_str = image_files[num].split("/")[-1].split(".")[0]
numGT = searchD(f_str, image_GT)
print("PAIRS-->", f_str)
print(image_GT[numGT])
# # filter for associated png annotations
# for root, _, files in os.walk(ANNOTATION_DIR):
class_id = 1 # SOS ayto allagh !!! isxyei gia to diko mou path !
# print ("File=", image_name)
# print (class_id)
# print (ROOT_DIR+'/masks/'+image_files[num].split("/")[-1])
# if "GT" in image_files[num].split("/")[-1]:
# # print(image)
# cv2.imwrite(ROOT_DIR+'/masks/'+image_files[num].split("/")[-1], image)
# os.remove(os.path.join(image_files[num]))
annotation_info = create_annotation(image_GT[numGT], image_id,
class_id)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
# segmentation_id = segmentation_id + 1
image_id = image_id + 1
json_name = 'annotdatasetMin.json'
json_path = '{}/annot/' + json_name
with open(json_path.format(ROOT_DIR), 'w') as output_json_file:
json.dump(coco_output, output_json_file)
def main():
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_id = 1
segmentation_id = 1
# filter for jpeg images
for img_root, _, img_files in os.walk(IMAGE_DIR):
image_files = filter_for_jpeg(img_root, img_files)
# go through each image
for image_filename in image_files:
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size)
coco_output["images"].append(image_info)
# filter for associated png annotations
for annotation_root, _, annotation_filenames in os.walk(ANNOTATION_DIR):
annotation_files = filter_for_annotations(annotation_root, annotation_filenames, image_filename)
# go through each associated annotation
for annotation_filename in annotation_files:
print(annotation_filename)
class_id = [x['id'] for x in CATEGORIES if x['name'] in annotation_filename][0]
category_info = {'id': class_id, 'is_crowd': 'crowd' in image_filename}
binary_mask = np.asarray(Image.open(annotation_filename).convert('1')).astype(np.uint8)
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id, image_id, category_info, binary_mask,
image.size, tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
with open(os.path.join(ROOT_DIR, JSON_FILENAME), 'w') as output_json_file:
json.dump(coco_output, output_json_file)
def main(config):
ROOT_DIR = config.ROOT_DIR_PYCOCOCREATOR
IMAGE_DIR =config.IMAGE_DIR
ANNOTATION_DIR = config.ANNOTATION_DIR
FILE_NAME = config.JSON_OUTPUT_FILENAME
CATEGORIES, coco_output = create_structure(config)
image_id = 1
segmentation_id = 1
# filter for jpeg images
for root, _, files in os.walk(IMAGE_DIR):
image_files = filter_for_jpeg(root, files)
# go through each image
for image_filename in image_files:
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(image_id, os.path.basename(image_filename), image.size)
coco_output["images"].append(image_info)
# filter for associated png annotations
for root, _, files in os.walk(ANNOTATION_DIR):
annotation_files = filter_for_annotations(root, files, image_filename)
# go through each associated annotation
for annotation_filename in annotation_files:
print(annotation_filename)
class_id = [x['id'] for x in CATEGORIES if x['name'] in annotation_filename][0]
category_info = {'id': class_id, 'is_crowd': 'crowd' in image_filename}
binary_mask = np.asarray(Image.open(annotation_filename)
.convert('1')).astype(np.uint8)
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id, image_id, category_info, binary_mask,
image.size, tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
with open('{}/{}'.format(ROOT_DIR, FILE_NAME), 'w') as output_json_file:
json.dump(coco_output, output_json_file)
def get_images(IMAGE_DIR, image_id):
images=[]
# filter for jpeg images
for root, _, files in os.walk(IMAGE_DIR):
image_files = filter_for_png(root, files)
image_files.sort() # ensure order
# go through each image
for image_filename in tqdm(image_files):
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size
)
images.append(image_info)
image_id +=1
return images,image_id
def mainfun(f_path,f2_path, bg_path, k, image_id, rangeN, class_idX, class_idY, coco_output, imageDir = IMAGE_DIR):
for i in range(rangeN):
image_filename = str(k) + ".jpg"
img = cv2.imread(random.choice(f_path))
print (img.shape[:2])
image_info = pycococreatortools.create_image_info(image_id, os.path.basename(image_filename), img.shape[:2])
coco_output["images"].append(image_info)
class_id = class_idX #image_filename.split("_")[-2][-1] # SOS ayto allagh !!! isxyei gia to diko mou path !
print ("File=", image_filename)
print ("class_id", class_id)
im_path= IMAGE_DIR + "/"+ image_filename
cv2.imwrite(im_path, img)
annotation_info = create_annotation(im_path, image_id, class_id)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
# segmentation_id = segmentation_id + 1
bgname = random.choice(bg_path)
imgBack = cv2.imread(bgname)
imFirst = replaceBG(img, imgBack)
if f2_path!=None:
img2 = cv2.imread(random.choice(f2_path))
cv2.imwrite(im_path, img2)
class_id = class_idY #image_filename.split("_")[-2][-1] # SOS this works for this !
print ("File=", image_filename)
print (class_id)
annotation_info = create_annotation(im_path, image_id, class_id)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
imFinal = replaceBG(img2, imFirst)
cv2.imwrite(im_path, imFinal)
image_id = image_id + 1
k -=1
else:
cv2.imwrite(im_path, imFirst)
image_id = image_id + 1
k -=1
return coco_output, k, image_id
def main():
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_id = 0
# segmentation_id = 1
# filter for jpeg images
for root, _, files in os.walk(IMAGE_DIR):
image_files = filter_for_jpeg(root, files)
# # print (IMAGE_DIR)
# # print (image_files)
# # go through each image
# print(image_files)
for image_filename in image_files:
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size)
print (image.size)
coco_output["images"].append(image_info)
# # filter for associated png annotations
# for root, _, files in os.walk(ANNOTATION_DIR):
class_id = image_filename.split("_")[-2][-1] # SOS ayto allagh !!! isxyei gia to diko mou path !
print ("File=", image_filename)
print (class_id)
annotation_info = create_annotation(image_filename, image_id, class_id)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
# segmentation_id = segmentation_id + 1
image_id = image_id + 1
json_name = '2clAluVal.json'
json_path = '{}/annot/'+ json_name
with open(json_path.format(ROOT_DIR), 'w') as output_json_file:
json.dump(coco_output, output_json_file)
def create_json_single_core(proc_id,
png_files,
inst_v_dir,
id_converter,
neighbor=False,
rle=False):
image_info = []
annotation_info = []
for idx, png_file in enumerate(png_files):
# iid = int(png_file.split('_')[0]+png_file.split('_')[3])
vid = int(png_file[:4])
iid = int(png_file[:9])
# print(iid)
# file_path = osp.join(INSTANCE_DIR, inst_v_dir, png_file)
file_path = osp.join(PANOPTIC_DIR, inst_v_dir, png_file)
insts = png2insts(file_path, id_converter, rle)
for inst in insts:
inst['image_id'] = iid
if inst['fcn_id'] <= 10: # cuz stuff_max = 10
print('Warning: wrong FCN id')
continue
inst['category_id'] = inst['fcn_id']
# inst['category_id'] = FCNID2TRAINID[inst['fcn_id']]
inst['width'] = W
inst['height'] = H
img_name = os.path.join(
inst_v_dir,
png_file.replace('final_mask',
'newImg8bit').replace('gtFine_color',
'leftImg8bit'))
im_info = pycococreatortools.create_image_info(iid, MODE, img_name,
(W, H))
image_info.append(im_info)
annotation_info.extend(insts)
return image_info, annotation_info
def worker_create_annotation_info(image_filename, image_id):
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(image_id, os.path.basename(image_filename), image.size)
#多进程计数器必须标明global
global ann_counter
# filter for associated png annotations
annotation_infos=[]
for root, _, files in os.walk(MASK_DIR):
annotation_files = filter_for_annotations(root, files, image_filename)
# go through each associated annotation
for annotation_filename in annotation_files:
class_id = [x['id'] for x in CATEGORIES if x['name'] in os.path.basename(annotation_filename)][0]
binary_mask = np.asarray(Image.open(annotation_filename).convert('1')).astype(np.uint8)
category_info = {'id': class_id, 'is_crowd': 0}
annotation_info = pycococreatortools.create_annotation_info(
ann_counter.increment(), image_id, category_info, binary_mask,
image.size, tolerance=2)
if annotation_info is not None:
annotation_infos.append(annotation_info)
return image_info, annotation_infos
def main():
EVAL_PUSH_NUMBER = -1
ROOT_DIR = sys.argv[1]
if len(sys.argv) == 3:
EVAL_PUSH_NUMBER = int(sys.argv[2])
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_id = 1
segmentation_id = 1
for trial_dir in os.listdir(ROOT_DIR):
if os.path.isdir(os.path.join(ROOT_DIR, trial_dir)):
print(trial_dir)
image_dir = os.path.join(ROOT_DIR, trial_dir, IMAGE_DIR)
annotation_dir = os.path.join(ROOT_DIR, trial_dir, ANNOTATION_DIR)
print("image dir", image_dir)
print("annotation dir", annotation_dir)
# filter for png images
for root, _, files in os.walk(image_dir):
files.sort()
if (EVAL_PUSH_NUMBER > len(files) - 1):
#files = files[len(files)-1:len(files)]
continue
elif (EVAL_PUSH_NUMBER >= 0):
files = files[EVAL_PUSH_NUMBER:EVAL_PUSH_NUMBER + 1]
image_files = filter_for_png(root, files)
image_files.sort()
#go through each image
for image_filename in image_files:
#print("image filename",image_filename)
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, relpath(image_filename, ROOT_DIR),
image.size)
coco_output["images"].append(image_info)
# filter for associated png annotations
for root, _, files in os.walk(annotation_dir):
files.sort()
annotation_files = filter_for_annotations(
root, files, image_filename)
if (len(annotation_files) == 0):
print("No annotations for", image_filename)
coco_output["images"].remove(image_info)
break
for annotation_filename in annotation_files:
print(annotation_filename)
class_id = 1
# Assign workspace class
if annotation_filename.find("_11.png") > -1:
class_id = 2
#print("class id",class_id)
category_info = {'id': class_id, 'is_crowd': 0}
binary_mask = cv2.imread(annotation_filename, -1)
print("Segmentation id", segmentation_id)
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id,
image_id,
category_info,
binary_mask,
image.size,
tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(
annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
basename = os.path.basename(ROOT_DIR)
basename = ''.join([i for i in basename if not i.isdigit()])
if not os.path.exists(ROOT_DIR + '/annotations'):
os.makedirs(ROOT_DIR + '/annotations')
outname = ROOT_DIR + '/annotations' + '/instances_push' + str(
EVAL_PUSH_NUMBER).zfill(2) + '.json'
with open(outname, 'w') as output_json_file:
json.dump(coco_output, output_json_file)
print("Saved json file", outname)
def main():
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": [],
}
image_id = 1
segmentation_id = 1
# filter for jpeg images
for root, _, files in os.walk(IMAGE_DIR):
# png_to_jpg(root, files)
image_files = filter_for_jpeg(root, files)
# go through each image
for image_filename in image_files:
print(image_filename)
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size)
coco_output["images"].append(image_info)
# filter for associated png annotations
for root, _, files in os.walk(ANNOTATION_DIR):
annotation_files = filter_for_annotations(
root, files, image_filename)
# go through each associated annotation
for annotation_filename in annotation_files:
# TODO: naive bug fix，限定於cat是按照複雜度排列，仍有可能出錯
class_id = [
x["id"] for x in CATEGORIES
if x["name"] in annotation_filename
][-1]
category_info = {
"id": class_id,
# "is_crowd": "crowd" in image_filename,
"is_crowd": True, # 強制讓segmentation用png->binary mask
}
binary_mask = png_to_binary_mask(annotation_filename)
annotation_info = pycococreatortools.create_annotation_info_without_mask(
segmentation_id,
image_id,
category_info,
binary_mask,
image.size,
# tolerance=10,
# tolerance=2,
)
# print(annotation_info)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
with open("{}/annotation.json".format(ROOT), "w") as output_json_file:
json.dump(coco_output, output_json_file)
def main():
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_id = 1
segmentation_id = 1
c = 0
# filter for jpeg images
for root, _, files in os.walk(IMAGE_DIR):
image_files = filter_for_jpeg(root, files)
# go through each image
for image_filename in image_files:
if 'mass' and 'mask' not in image_filename:
# print(image_filename)
print(
str(c + 1) + '/' +
str(sum([len(files)
for _, _, files in os.walk(IMAGE_DIR)])))
c += 1
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size)
coco_output["images"].append(image_info)
# filter for associated png annotations
for root_ann, _, ann in os.walk(ANNOTATION_DIR):
annotation_files = filter_for_annotations(
root_ann, ann, image_filename)
# go through each associated annotation
for annotation_filename in annotation_files:
# print(annotation_filename)
class_id = 0
if 'mass' or 'mask' in annotation_filename:
class_id = 1
category_info = {
'id': class_id,
'is_crowd': 'crowd' in image_filename
}
binary_mask = np.asarray(
Image.open(annotation_filename).convert(
'1')).astype(np.uint8)
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id,
image_id,
category_info,
binary_mask,
image.size,
tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
else:
print('No image files')
continue
with open(os.path.join(ROOT_DIR, 'instances_shape_test2019.json'),
'w') as output_json_file:
json.dump(coco_output, output_json_file)
def main():
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": [],
}
image_id = 1
segmentation_id = 1
# filter for jpeg images
for root, _, files in os.walk(IMAGE_DIR):
image_files = filter_for_png(root, files)
image_files.sort() # ensure order
# go through each image
for image_filename in tqdm(image_files):
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size)
coco_output["images"].append(image_info)
# filter for associated png annotations
for root, _, files in os.walk(ANNOTATION_DIR):
annotation_files = filter_for_annotations(
root, files, image_filename)
# go through each associated annotation
for annotation_filename in annotation_files:
class_id = int(
re.search(r'\d(?=_inst\d.png)',
annotation_filename).group())
category_info = {
"id": class_id,
"is_crowd": "crowd" in image_filename,
}
binary_mask = np.asarray(
Image.open(annotation_filename)).astype(np.uint8)
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id,
image_id,
category_info,
binary_mask,
image.size,
tolerance=2,
)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
with open(
"{0}/{1}_{2}.json".format(SAVE_DIR, args.json_prefix, GROUP_NAME),
"w") as output_json_file:
json.dump(coco_output, output_json_file, indent=4)
def create_json_direct(images_path, new_path):
'''
:param images_path:
:param new_path:
:return:
'''
def get_mask_path(im_path):
image_name = os.path.basename(im_path)
sub_dir_path = os.path.dirname(os.path.dirname(im_path))
mask_dir_name = image_name.split('.')[0] + "_outlines"
mask_dir_path = os.path.join(sub_dir_path, "crops", mask_dir_name)
mask_image_name = os.listdir(mask_dir_path)
mask_image_name = [
i for i in mask_image_name if len(i.split('_')) == 3
]
mask_image_path = [
os.path.join(mask_dir_path, i) for i in mask_image_name
]
return mask_image_path
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
train_images_path = os.path.join(new_path, "train2020_scale")
if not os.path.exists(train_images_path):
os.mkdir(train_images_path)
train_anno_path = os.path.join(new_path, "annotations",
"instances_cell_train2020_scale.json")
train_image_bad_path = os.path.join(new_path, "bad_ann")
images_num = len(images_path)
image_id = 1
segmentation_id = 1
print("Total image len:{}".format(images_num))
for im_path in images_path:
# new_image_name is the new name of this image in train2020_sacle and coco json file
# './new_training/76-9-10052-crop2/xy-images/crop2-Composite-10051.tif'
print("=====================================")
print(im_path)
dir_name = im_path.split('/')[-3]
im = os.path.basename(im_path).split('.')[0]
im = '-'.join(im.split('-')[-2:])
new_image_name = dir_name + "-" + im + ".png"
print(new_image_name)
# read image and save to new dir
# image.size 是 w,h
image = Image.open(im_path)
image_info = pycococreatortools.create_image_info(
image_id, new_image_name, image.size)
coco_output["images"].append(image_info)
print("image_info:\n{}".format(image_info))
# save to new dir
new_image_path = os.path.join(train_images_path, new_image_name)
image.save(new_image_path)
print("Save as:{}".format(new_image_path))
# find this image`s masks and make boxes & segmentation info
masks_path = get_mask_path(im_path)
print("Masks num of image:{}".format(len(masks_path)))
for c in range(len(masks_path)):
ms_path = masks_path[c]
# ms_path is one of the masks path for this image
binary_mask = np.asarray(Image.open(ms_path).convert("L")).astype(
np.uint8)
binary_mask[binary_mask > 0] = 1
class_id = 1
category_info = {'id': class_id, 'is_crowd': 0}
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id,
image_id,
category_info,
binary_mask,
image.size,
tolerance=2)
print("image size:{} {} mask size:{}".format(
image.size[1], image.size[0], binary_mask.shape))
print("Annotation info:{}".format(annotation_info))
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
else:
# coco_output["annotations"].append(annotation_info)
print("There is an bad binary mask in :{}-{}".format(
new_image_name, c))
save_bad_ann(train_image_bad_path, new_image_path, binary_mask,
segmentation_id)
segmentation_id += 1
image_id += 1
with open(train_anno_path, 'w') as output_json_file:
# json.dump(coco_output, output_json_file)
json.dump(coco_output, output_json_file, indent=4)
def __call__(self, img_desc):
img_dir, img_id, lbl_cat = img_desc
img_unique_id = counter.increment()
coco_ann = []
# Load the annotation
with Image.open(
path.join(self.msk_base_dir, img_dir,
img_id + "_" + lbl_cat + _INSTANCE_EXT)) as lbl_img:
lbl = np.array(lbl_img)
lbl_size = lbl_img.size
ids = np.unique(lbl)
# Compress the labels and compute cat
lbl_out = np.ones(lbl.shape, np.uint8) * 255
cat = [255]
iscrowd = [0]
for city_id in ids:
if city_id < 1000:
# Stuff or group
cls_i = city_id
iscrowd_i = cs_labels[cls_i].hasInstances
else:
# Instance
cls_i = city_id // 1000
iscrowd_i = False
# If it's a void class just skip it
if cs_labels[cls_i].trainId == 255 or cs_labels[
cls_i].trainId == -1:
continue
# Extract all necessary information
iss_class_id = cs_labels[cls_i].trainId
mask_i = lbl == city_id
lbl_out[mask_i] = iss_class_id
# Compute COCO detection format annotation
if cs_labels[cls_i].hasInstances:
category_info = {"id": iss_class_id, "is_crowd": iscrowd_i}
coco_ann_i = pct.create_annotation_info(counter.increment(),
img_unique_id,
category_info,
mask_i,
lbl_size,
tolerance=2)
if coco_ann_i is not None:
coco_ann.append(coco_ann_i)
# COCO detection format image annotation
coco_img = pct.create_image_info(img_unique_id, img_id + ".png",
lbl_size)
# Write output
Image.fromarray(lbl_out).save(path.join(self.msk_dir, img_id + ".png"))
shutil.copy(path.join(self.img_base_dir, img_dir, img_id + _IMAGE_EXT),
path.join(self.img_dir, img_id + ".png"))
return coco_img, coco_ann
def convert(self, train_type):
mask_dir = self.ROOT_DIR + '/Part_Masks/{}'.format(train_type)
image_dir = self.ROOT_DIR + '/Images/{}'.format(train_type)
coco_output = {
"info": self.INFO,
"licenses": self.LICENSES,
"categories": self.CATEGORIES,
"images": [],
"annotations": []
}
image_id = 1
segmentation_id = 1
# filter for jpeg images
for root, _, files in os.walk(image_dir):
image_files = self.filter_for_jpeg(root, files)
# go through each image
for index, image_filename in enumerate(tqdm(image_files)):
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size)
coco_output["images"].append(image_info)
# filter for associated png annotations
for root, _, files in os.walk(mask_dir):
annotation_files = self.filter_for_annotations(
root, files, image_filename)
# go through each associated annotation
for annotation_filename in annotation_files:
# print(annotation_filename)
class_id = [
x['id'] for x in self.CATEGORIES
if x['name'] in annotation_filename
][0]
category_info = {
'id': class_id,
'is_crowd': 'crowd' in image_filename
}
binary_mask = np.asarray(
Image.open(annotation_filename).convert(
'1')).astype(np.uint8)
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id,
image_id,
category_info,
binary_mask,
image.size,
tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
with open('{}/{}.json'.format(self.ROOT_DIR, train_type),
'w') as output_json_file:
json.dump(coco_output, output_json_file)
def json_generate():
car = 0
pedestrian = 0
files = os.listdir(IMAGE_SAVE_DIR)
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_id = 1
segmentation_id = 1
# go through each image
for image_filename in files:
image_name = image_filename.split('.')[0]
image_path = os.path.join(IMAGE_SAVE_DIR, image_filename)
image = Image.open(image_path)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size)
coco_output["images"].append(image_info)
print(image_filename)
annotation_sub_path = os.path.join(INSTANCE_DIR, image_name)
ann_files = os.listdir(annotation_sub_path)
if len(ann_files) == 0:
print("ao avaliable annotation")
continue
else:
for annotation_filename in ann_files:
annotation_path = os.path.join(annotation_sub_path,
annotation_filename)
for x in CATEGORIES:
if x['name'] in annotation_filename:
class_id = x['id']
break
if class_id == 1:
car += 1
elif class_id == 2:
pedestrian += 1
else:
print('illegal class id')
category_info = {
'id': class_id,
'is_crowd': 'crowd' in image_filename
}
binary_mask = np.asarray(
Image.open(annotation_path).convert('1')).astype(np.uint8)
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id,
image_id,
category_info,
binary_mask,
image.size,
tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
print(image_id)
with open('{}/val.json'.format(ROOT_DIR), 'w') as output_json_file:
json.dump(coco_output, output_json_file)
print(car, pedestrian)
def main():
""" Main entry point of the app """
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
segmentation_id = 1
image_id = 1
for root, directories, files in os.walk(IMAGE_DIR):
file_types = ['*.jpeg', '*.jpg']
file_types = r'|'.join([fnmatch.translate(x) for x in file_types])
files = [os.path.join(root, f) for f in files]
files = [f for f in files if re.match(file_types, f)]
# go through each image
for i, filename in enumerate(files):
print(filename)
parent_directory = root.split(os.path.sep)[-1]
basename_no_extension = os.path.splitext(
os.path.basename(filename))[0]
image = Image.open(filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(filename), image.size)
#print(image_info)
coco_output["images"].append(image_info)
# go through each associated annotation
for root, directories, files in os.walk(ANNOTATION_DIR):
file_types = ['*.png']
file_types = r'|'.join(
[fnmatch.translate(x) for x in file_types])
file_name_prefix = basename_no_extension + '.*'
files = [os.path.join(root, f) for f in files]
files = [f for f in files if re.match(file_types, f)]
files = [
f for f in files
if re.match(file_name_prefix,
os.path.splitext(os.path.basename(f))[0])
]
for filename in files:
parent_directory = root.split(os.path.sep)[-1]
basename_no_extension = os.path.splitext(
os.path.basename(filename))[0]
annotation_array = np.array(Image.open(filename))
# load annotation information
annotation_metadata = get_metadata(filename)
print(annotation_metadata)
annotation_data = AnnotationData(annotation_array,
annotation_metadata)
object_classes = annotation_data.get_classes()
#print(object_classes)
# go through each class
for j, object_class in enumerate(object_classes):
if ANNOTATOR_CATEGORIES.get(object_class) == None:
print("missing: {}".format(object_class))
continue
# go through each object
for object_instance in range(
object_classes[object_class]):
object_mask = annotation_data.get_mask(
object_class, object_instance)
if object_mask is not None:
object_mask = object_mask.astype(np.uint8)
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id,
image_id,
ANNOTATOR_CATEGORIES.get(object_class),
object_mask,
image.size,
tolerance=2)
print(annotation_info)
if annotation_info is not None:
coco_output["annotations"].append(
annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
with open('{}/coco.json'.format(DATA_DIR), 'w') as output_json_file:
json.dump(coco_output, output_json_file)
def main():
image_id = 1
segmentation_id = 1
VIDEOS =[]
ANNOTATIONS =[]
IMAGES_VAL =[]
IMAGES_TRAIN =[]
ANNOTATIONS_VAL =[]
ANNOTATIONS_TRAIN =[]
root_dir = IMAGE_DIR
train, val = data_split(FOLD_NUM)
# filter for jpeg images
for dir in os.listdir(root_dir):
print(dir)
path = os.path.join(root_dir,dir)+"/left_frames/"
ANNOTATION_DIR = os.path.join(root_dir,dir)+"/ground_truth/"
#print(path)
# filter for jpeg images
for root, _, files in os.walk(path):
image_files = [os.path.join(path,f) for f in files]#filter_for_png(root, files)
image_files.sort() # ensure order
#print(image_files)
# go through each image
for image_filename in image_files: #tqdm(image_files):
image = Image.open(image_filename)
#print(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size
)
image_info["file_name"]=os.path.join(dir,image_info["file_name"])
# filter for associated png annotations
annotation_files = filter_for_annotations(
ANNOTATION_DIR,image_filename
)
#print("here",annotation_files)
if len(annotation_files):
if dir not in train:
IMAGES_VAL.append(image_info)
else:
IMAGES_TRAIN.append(image_info)
image_id +=1
# go through each associated annotation
for annotation_filename in annotation_files:
label,class_id = get_label(annotation_filename.split("/")[-2])
if not class_id:
print(annotation_filename)
print(label,class_id)
category_info = {
"id": class_id
}
binary_mask = cv2.imread(annotation_filename,0).astype(np.uint8)
area,bbox,segmentation = create_annotation_info(binary_mask,None)
if area:
annotation_info= {
"id": segmentation_id,
"image_id": image_id,
"category_id": class_id,
"iscrowd":None,
"area": area,
"bbox": bbox,
"segmentation": segmentation,
"width": binary_mask.shape[1],
"height": binary_mask.shape[0],
}
segmentation_id+=1
if dir not in train:
ANNOTATIONS_VAL.append(annotation_info)
else:
ANNOTATIONS_TRAIN.append(annotation_info)
coco_output_train = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": IMAGES_TRAIN,
"annotations": ANNOTATIONS_TRAIN,
}
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": IMAGES_VAL,
"annotations": ANNOTATIONS_VAL,
}
with open( "instances_train_sub.json", "w") as output_json_file:
json.dump(coco_output_train, output_json_file, indent=4)
with open( "instances_val_sub.json", "w") as output_json_file:
json.dump(coco_output, output_json_file, indent=4)
with open(template_path) as f:
data_val = json.load(f)
# Loop all files and annotate them
print(str(num_images) + " images found")
annotation_id = 0
num_images = len([name for name in os.listdir(IMAGE_DIR) if os.path.isfile(os.path.join(IMAGE_DIR, name))])
for i, image_file in enumerate(sorted(os.listdir(IMAGE_DIR))):
print("{}: working on image {}".format(i, image_file))
image = skimage.io.imread(os.path.join(IMAGE_DIR, image_file))
a, b = image.shape[0:2]
image_size = b, a
image_id = int(image_file.split('_')[-1][:-4])
img_data = pycococreatortools.create_image_info(image_id, image_file, image_size)
if i < num_images * DATA_SPLIT:
data_train["images"].append(img_data)
else:
data_val["images"].append(img_data)
# Run detection
try:
results = model.detect([image])
# Continue if no detection was made
except IndexError:
print("No detection found")
annotation_errors[image_file] = "No detection found"
continue
except:
def main():
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_id = 1
segmentation_id = 1
# filter for jpeg images
for root, _, files in os.walk(IMAGE_DIR):
image_files = filter_for_jpeg(root, files)
# go through each image
for image_filename in image_files:
image = Image.open(image_filename)
image_info = pycococreatortools.create_image_info(
image_id, os.path.basename(image_filename), image.size)
coco_output["images"].append(image_info)
# filter for associated png annotations
for root, _, files in os.walk(ANNOTATION_DIR):
annotation_files = filter_for_annotations(root, files, image_filename)
# go through each associated annotation
for annotation_filename in annotation_files:
print(annotation_filename)
if 'c01' in annotation_filename:
class_id = 1
elif 'c02' in annotation_filename:
class_id = 2
elif 'c03' in annotation_filename:
class_id = 3
elif 'c04' in annotation_filename:
class_id = 4
elif 'c05' in annotation_filename:
class_id = 5
elif 'c06' in annotation_filename:
class_id = 6
elif 'c07' in annotation_filename:
class_id = 7
elif 'c08' in annotation_filename:
class_id = 8
elif 'c09' in annotation_filename:
class_id = 9
elif 'c10' in annotation_filename:
class_id = 10
elif 'c11' in annotation_filename:
class_id = 11
elif 'c12' in annotation_filename:
class_id = 12
elif 'c13' in annotation_filename:
class_id = 13
elif 'c14' in annotation_filename:
class_id = 14
elif 'c15' in annotation_filename:
class_id = 15
elif 'c16' in annotation_filename:
class_id = 16
elif 'c17' in annotation_filename:
class_id = 17
elif 'c18' in annotation_filename:
class_id = 18
elif 'c19' in annotation_filename:
class_id = 19
elif 'c20' in annotation_filename:
class_id = 20
elif 'c21' in annotation_filename:
class_id = 21
elif 'c22' in annotation_filename:
class_id = 22
elif 'c23' in annotation_filename:
class_id = 23
else:
class_id = 24
category_info = {'id': class_id, 'is_crowd': 'crowd' in image_filename}
binary_mask = np.asarray(Image.open(annotation_filename)
.convert('1')).astype(np.uint8)
annotation_info = pycococreatortools.create_annotation_info(
segmentation_id, image_id, category_info, binary_mask,
image.size, tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
segmentation_id = segmentation_id + 1
image_id = image_id + 1
with open('{0}/MFISH_chromosomes_{1}2018.json'.format(ANNOTATION_DIR, SET_NAME), 'w') as output_json_file:
json.dump(coco_output, output_json_file)
def main():
CLASS_ID = 1
CATEGORIES = \
[{
'id': CLASS_ID,
'name': 'jenga_block',
'supercategory': 'shape',
}]
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"images": [],
"annotations": []
}
image_global_id = 1
segmentation_global_id = 1
# filter for jpeg images
img_seg_files_dict = filter_for_jpeg()
# print(img_seg_files_dict)
img_size = (224, 224)
for img_file_path in list(img_seg_files_dict.keys()):
image_info = pycococreatortools.create_image_info(
image_global_id, img_file_path, img_size)
mask_valid = False
for seg_file_path in img_seg_files_dict[img_file_path]:
# Crowd has to be 0 or else no anno get loaded in maskrcnn
category_info = {'id': CLASS_ID, 'is_crowd': 0}
# print(seg_file_path)
binary_mask = np.asarray(Image.open(seg_file_path))
if np.count_nonzero(binary_mask) > 0:
annotation_info = pycococreatortools.create_annotation_info(
segmentation_global_id,
image_global_id,
category_info,
binary_mask,
img_size,
tolerance=2)
if annotation_info is not None:
coco_output["annotations"].append(annotation_info)
mask_valid = True
segmentation_global_id += 1
if mask_valid == True:
# pss
coco_output["images"].append(image_info)
else:
print("No valid mask")
if image_global_id % 100 == 0:
print("Image {}/{}".format(image_global_id,
len(list(img_seg_files_dict.keys()))))
image_global_id = image_global_id + 1
with open('{}/{}.json'.format(ROOT_OUTDIR, OUTFILE_NAME),
'w') as output_json_file:
json.dump(coco_output, output_json_file)
def main():
# object_settings_file = Path(os.path.join(IMAGE_DIR_LIST[0], "_object_settings.json"))
# camera_settings_file = Path(os.path.join(IMAGE_DIR_LIST[0], "_camera_settings.json"))
if object_settings_file.is_file():
with open(object_settings_file) as file:
object_settings_data = json.load(file)
if SELECTED_OBJECTS is None:
CLASSES = object_settings_data['exported_object_classes']
else:
CLASSES = SELECTED_OBJECTS
CATEGORIES = [{
'id': i,
'name': CLASSES[i].replace('_16k', '').replace('_16K', ''),
'supercategory': 'shape',
} for i in range(0,len(CLASSES))]
FIXED_TRANSFORMS = {}
for i in range(0,len(object_settings_data['exported_object_classes'])):
class_name = object_settings_data['exported_objects'][i]['class']
transform = object_settings_data['exported_objects'][i]['fixed_model_transform']
if class_name in CLASSES:
class_name = class_name.replace('_16k', '').replace('_16K', '')
FIXED_TRANSFORMS[class_name] = transform
# print(FIXED_TRANSFORMS)
# SEGMENTATION_DATA = object_settings_data['exported_objects']
else:
raise Exception("Object settings file not found")
if camera_settings_file.is_file():
with open(camera_settings_file) as file:
camera_settings_data = json.load(file)
CAMERA_INTRINSICS = camera_settings_data['camera_settings'][0]['intrinsic_settings']
else:
raise Exception("Camera settings file not found")
VIEWPOINTS = [viewpoints_xyz[i].tolist() for i in range(0, len(viewpoints_xyz))]
INPLANE_ROTATIONS = [inplane_rot_angles[i] for i in range(0, len(inplane_rot_angles))]
coco_output = {
"info": INFO,
"licenses": LICENSES,
"categories": CATEGORIES,
"viewpoints" : VIEWPOINTS,
"inplane_rotations" : INPLANE_ROTATIONS,
"camera_intrinsic_settings": CAMERA_INTRINSICS,
"fixed_transforms": FIXED_TRANSFORMS,
"images": [],
"annotations": []
}
image_global_id = 1
segmentation_global_id = 1
# filter for jpeg images
for IMAGE_DIR_T in IMAGE_DIR_LIST:
# for root, _, files in os.walk(IMAGE_DIR):
for SCENE in SCENES:
if IMAGE_DIR_T == "":
IMAGE_DIR = os.path.join(ROOT_DIR, SCENE)
else:
IMAGE_DIR = os.path.join(ROOT_DIR, IMAGE_DIR_T, SCENE)
all_dir_files = os.listdir(IMAGE_DIR)
image_files = filter_for_jpeg(IMAGE_DIR, all_dir_files)
# dir_name = os.path.basename(IMAGE_DIR)
SEGMENTATION_DATA = get_segmentation_data_for_scene(IMAGE_DIR)
print(SEGMENTATION_DATA)
# go through each image
for ii in trange(len(image_files)):
image_filename = image_files[ii]
if IMAGE_DIR_T == "":
image_out_filename = os.path.join(SCENE, os.path.basename(image_filename))
else:
image_out_filename = os.path.join(IMAGE_DIR_T, SCENE, os.path.basename(image_filename))
img_size = (960,540)
image_info = pycococreatortools.create_image_info(
image_global_id, image_out_filename, img_size
)
# plt.figure()
# skimage.io.imshow(skimage.io.imread(image_filename))
# plt.show()
# filter for associated png annotations
# for root, _, files in os.walk(IMAGE_DIR):
segmentation_image_files = filter_for_annotations(IMAGE_DIR, all_dir_files, image_filename)
label_files = filter_for_labels(IMAGE_DIR, all_dir_files, image_filename)
boxes = []
labels = []
segmentation_ids = []
label_filename = label_files[0]
# go through each associated json file containing objects data
# for label_filename in label_files:
# print("File %d - %s"% (image_global_id, label_filename))
my_file = Path(label_filename)
segmentation_image = skimage.io.imread(segmentation_image_files[0])
# print("File %d - %s"% (image_global_id, segmentation_image_files[0]))
if my_file.is_file():
with open(label_filename) as file:
label_data = json.load(file)
# all_objects_yaw_only = True
for i in range(0, len(label_data['objects'])):
class_name = label_data['objects'][i]['class']
if class_name not in SELECTED_OBJECTS:
continue
# print(class_name)
class_bounding_box = label_data['objects'][i]['bounding_box']
quat = label_data['objects'][i]['quaternion_xyzw']
angles = RT_transform.quat2euler(get_wxyz_quaternion(quat))
# angles = RT_transform.quat2euler(get_wxyz_quaternion(quat), 'syxz')
# angles = apply_angle_symmetry(angles, SYMMETRY_INFO[class_name])
# This function gives angles with this convention of euler - https://en.wikipedia.org/wiki/Euler_angles#Signs_and_ranges (geometric definition)
# if np.isclose(angles[1], 0):
# print("Test")
theta, phi = euler2sphere(angles[1], angles[0])
actual_angles = np.array([1, theta, phi])
xyz_coord = sphere2cart(1, theta, phi)
viewpoint_id = find_viewpoint_id(viewpoints_xyz, xyz_coord)
r_xyz = get_viewpoint_from_id(viewpoints_xyz, viewpoint_id)
recovered_angles = np.array(cart2sphere(r_xyz[0], r_xyz[1], r_xyz[2]))
inplane_rotation_id = find_inplane_rotation_id(inplane_rot_angles, angles[2])
# inplate_rotation_angle = get_inplane_rotation_from_id(INPLANE_ROTATIONS, inplane_rotation_id)
if np.all(np.isclose(actual_angles, recovered_angles, atol=0.4)) == False:
print("Mismatch in : {}".format(label_filename))
print("sphere2cart angles : {}".format(actual_angles))
print("cart2sphere angles : {}".format(recovered_angles))
# elif np.all(np.isclose(actual_angles, recovered_angles, atol=0.4)) == True:
# print("Match")
# print(inplate_rotation_angle)
class_label = [x['id'] for x in CATEGORIES if x['name'] in class_name][0]
segmentation_id = [x['segmentation_class_id'] for x in SEGMENTATION_DATA if x['class'] in class_name][0]
boxes.append(class_bounding_box['top_left'] + class_bounding_box['bottom_right'])
labels.append(class_label)
segmentation_ids.append([x['segmentation_class_id'] for x in SEGMENTATION_DATA if x['class'] in class_name][0])
# Create binary masks from segmentation image for every object
# for segmentation_image_file in segmentation_image_files:
# segmentation_image = skimage.io.imread(segmentation_image_file)
binary_mask = np.copy(segmentation_image)
binary_mask[binary_mask != segmentation_id] = 0
binary_mask[binary_mask == segmentation_id] = 1
# skimage.io.imshow(binary_mask, cmap=plt.cm.gray)
# plt.show()
# TODO : check if its actually a crowd in case of multiple instances of one object type
# class_label = [x['class'] for x in SEGMENTATION_DATA if x['segmentation_class_id'] in segmentation_id][0]
category_info = {'id': class_label, 'is_crowd': 0}
annotation_info = pycococreatortools.create_annotation_info(
segmentation_global_id, image_global_id, category_info, binary_mask,
img_size, tolerance=2)
# print(annotation_info)
if annotation_info is not None:
annotation_info['viewpoint_id'] = int(viewpoint_id)
annotation_info['inplane_rotation_id'] = int(inplane_rotation_id)
annotation_info['camera_pose'] = label_data['camera_data']
annotation_info['location'] = label_data['objects'][i]['location']
annotation_info['quaternion_xyzw'] = quat
coco_output["annotations"].append(annotation_info)
coco_output["images"].append(image_info)
else:
tqdm.write("File %s doesn't have boxes or labels in json file" % image_filename)
segmentation_global_id = segmentation_global_id + 1
else:
tqdm.write("File %s doesn't have a label file" % image_filename)
image_global_id = image_global_id + 1
with open('{}/{}.json'.format(ROOT_DIR, OUTFILE_NAME), 'w') as output_json_file:
json.dump(coco_output, output_json_file)
