def classify(rgb_img):
config = CocoConfig()
model = modellib.MaskRCNN(mode="inference",
model_dir="./models/",
config=config)
model.load_weights(filepath="./models/mask_rcnn_moles_0030.h5",
by_name=True)
class_names = ['Salmonella', 'Yersinia', 'Shigella', 'Negativo']
r = model.detect([rgb_img])[0]
image, caption = visualize.display_instances(rgb_img, r['rois'],
r['masks'], r['class_ids'],
class_names, r['scores'])
K.clear_session()
image = PIL.Image.fromarray(image, 'RGB')
words = caption.split()
filename = str(uuid.uuid4())
filepath = "./static/img/test/{}.jpg".format(filename)
image.save(filepath)
return json.dumps([{
"caption": words[0],
"percent": words[1],
"image_path": filepath
}],
cls=NumpyEncoder)
def maskrandommiddle(level_str, fig_name,rando):
# game = 0;
# while game < 100:
# Load a random image from the images folder
level = 2
with graph.as_default():
if rando == 0:
file_names = os.path.join("D:\FYP\myfyp\Mask\images\middle", fig_name)
image = skimage.io.imread(os.path.join("D:\FYP\myfyp\Mask\images\middle", file_names))
else:
file_names = next(os.walk("D:\FYP\myfyp\Mask\images\middle"))[2]
image = skimage.io.imread(os.path.join("D:\FYP\myfyp\Mask\images\middle", random.choice(file_names)))
# Run detection
results = model.detect([image], verbose=1)
# Visualize results
r = results[0]
label_list = visualize.display_instances(level, image, r['rois'], r['masks'], r['class_ids'],
class_names, r['scores'])
return label_list[0], label_list[1], label_list[2]
'input_image_meta:0': image_metas.tolist()[0],
'input_anchors:0': anchors.tolist()[0]
}]
})
res = requests.post('http://86.56.135.139:31967/v1/models/skyn:predict',
data=data,
headers=headers)
decoded = json.loads(res.content.decode("utf-8"))['predictions'][0]
detections = np.array([decoded['mrcnn_detection/Reshape_1:0']])
mrcnn_mask = np.array([decoded['mrcnn_mask/Reshape_1:0']])
# Process detections
results = []
for i, image in enumerate(images):
final_rois, final_class_ids, final_scores, final_masks =\
model.unmold_detections(detections[i], mrcnn_mask[i],
image.shape, molded_images[i].shape,
windows[i])
results.append({
"rois": final_rois,
"class_ids": final_class_ids,
"scores": final_scores,
"masks": final_masks,
})
r = results[0]
print(r['scores'])
visualize.display_instances(img, r['rois'], r['masks'], r['class_ids'],
class_names, r['scores'])
mask = cv2.resize(mask, (128, 128))
if mask is not None:
print("*****Done*****")
info = Metadata(data["meta"], data["dataset"], img, mask)
train_data = info
# processing the data
dataset_train = MoleDataset()
dataset_train.load_shapes(train_data, config.IMAGE_SHAPE[0],
config.IMAGE_SHAPE[1])
dataset_train.prepare()
image = dataset_train.load_image(0)
mask, class_ids = dataset_train.load_mask(0)
visualize.test_display_image(image, mask, class_ids,
dataset_train.class_names)
r = model.detect([image])[0]
c = visualize.display_instances(image,
r['rois'],
r['masks'],
r['class_ids'],
class_names,
r['scores'],
figsize=(7, 5))
print("++++++++++DONE++++++++++++")
print("complete")
# Test on a random image
for i in range(10):
image_id = np.random.choice(dataset_val.image_ids)
original_image, image_meta, gt_class_id, gt_bbox, gt_mask =\
modellib.load_image_gt(dataset_val, inference_config,
image_id, use_mini_mask=False)
log("original_image", original_image)
log("image_meta", image_meta)
log("gt_class_id", gt_class_id)
log("gt_bbox", gt_bbox)
log("gt_mask", gt_mask)
visualize.display_instances(original_image,
gt_bbox,
gt_mask,
gt_class_id,
dataset_train.class_names,
figsize=(8, 8))
r = model.detect([original_image], verbose=1)[0]
def get_ax(rows=1, cols=1, size=8):
_, ax = plt.subplots(rows, cols, figsize=(size * cols, size * rows))
return ax
visualize.display_instances(original_image,
r['rois'],
r['masks'],
r['class_ids'],
dataset_val.class_names,
r['scores'],