def main():
print("Initializing ...")
working_directory=sys.argv[1]
if working_directory[-1]=="/":
working_directory=working_directory[0:-1]
source_csv=working_directory+"/input.csv"
csv_exportname=working_directory+"/output.csv"
pixel_range=(0,255)
mask_threshold=255/2
full_data=[["id","temp","raw","x","y"]]
i=1
with open(source_csv) as csv_file:
csv_reader=csv.DictReader(csv_file)
for frame in csv_reader:
# say what we're doing
print("Starting image "+str(i), end="")
sys.stdout.flush()
# process each entry in csv
# csv format: id, image name, mask name, max temp, min temp, crop details (2 points)
image_filename=working_directory+"/"+frame["image_name"]
mask_filename=working_directory+"/"+frame["mask_name"]
temp_range=(float(frame["max_temp"]),float(frame["min_temp"]))
cropsize=(int(frame["topleft_x"]), int(frame["topleft_y"]),
int(frame["bottomright_x"])-int(frame["topleft_x"]),
int(frame["bottomright_y"])-int(frame["topleft_y"]))
start_coords=(int(frame["topleft_x"]), int(frame["topleft_y"]))
id=frame["id"]
# load thermal image
main_image=ndimage.imread(image_filename,flatten=True)
cropped_image=imaging.crop(cropsize,main_image)
# process mask to boolean
mask_image=ndimage.imread(mask_filename,flatten=True)
cropped_mask=imaging.crop(cropsize,mask_image)
cropped_mask=cropped_mask[:,:] < mask_threshold
if not imaging.verify_mask(main_image, mask_image):
raise Exception("Mask does not fit image")
# make the dataset for this image
frame_data=thermal.thermal_image_to_dataset(cropped_image, temp_range, pixel_range, id, cropped_mask, start_coords)
full_data.extend(frame_data)
i+=1
print("done!")
# save the dataset
print("Saving csv ...")
with open(csv_exportname, mode="w", newline="") as file:
write_obj=csv.writer(file)
write_obj.writerows(full_data)
print("Completed!")
def test_crop3(self):
test_image = np.array([[1, 2], [3, 4]])
test_params = (0, 0, 2, 1)
expected_cropped_image = np.array([[1]])
actual_cropped_image = imaging.crop(test_params, test_image)
result = np.array_equal(actual_cropped_image, expected_cropped_image)
self.assertFalse(result)
def test_crop1(self):
test_image = np.array([[1, 2], [3, 4]])
test_params = (1, 1, 1, 1)
expected_cropped_image = np.array([[4]])
actual_cropped_image = imaging.crop(test_params, test_image)
result = np.array_equal(actual_cropped_image, expected_cropped_image)
self.assertTrue(result)
def main():
image_filename = sys.argv[1]
csv_exportname = sys.argv[2]
cropsize = (10, 30, 290, 180) #x offset, y offset, x size, y size
temp_range = (17.0, 12.9)
pixel_range = (0, 255)
mask_threshold = 255 / 2
# load thermal image
main_image = ndimage.imread(image_filename, flatten=True)
if len(sys.argv) == 4:
mask_filename = sys.argv[3]
mask_image = ndimage.imread(mask_filename, flatten=True)
cropped_mask = imaging.crop(cropsize, mask_image)
cropped_mask = cropped_mask[:, :] < mask_threshold
if not imaging.verify_mask(main_image, mask_image):
raise Exception("Mask does not fit image")
if config.debug:
plt.imshow(cropped_mask)
plt.show()
else:
cropped_mask = None
cropped_image = imaging.crop(cropsize, main_image)
if (config.debug):
plt.imshow(cropped_image)
plt.show()
test_datset = thermal.thermal_image_to_dataset(cropped_image, temp_range,
pixel_range, "1",
cropped_mask)
with open(csv_exportname, mode="w", newline="") as file:
write_obj = csv.writer(file)
write_obj.writerows(test_datset)
def main():
print("Initializing ...")
if len(sys.argv) > 1: # check if there's a folder argument
gui_enabled = False
working_directory = sys.argv[1]
if working_directory[-1] == "/":
working_directory = working_directory[0:-1]
source_csv = working_directory + "/input_batch.csv"
csv_exportname = working_directory + "/output.csv"
else:
gui_enabled = True
working_directory = easygui.diropenbox("Select image folder")
source_csv = working_directory + "/input_batch.csv"
csv_exportname = easygui.filesavebox("Save resulting CSV")
pixel_range = (0, 255)
mask_threshold = 255 / 2
full_data = [["id", "temp", "raw", "x", "y"]]
i = 1
with open(source_csv) as csv_file:
csv_reader = csv.DictReader(csv_file)
for frame in csv_reader:
# say what we're doing
print("Starting image " + str(i) + " (" + frame["id"] + ")",
end="")
sys.stdout.flush()
# process each entry in csv
# csv format: id, image name, mask name, max temp, min temp,
# crop details (2 x-y coordinates)
image_filename = working_directory + "/" + frame["image_name"]
mask_filename = working_directory + "/" + frame["mask_name"]
temp_range = (float(frame["max_temp"]), float(frame["min_temp"]))
id = frame["id"]
# load thermal image
main_image = ndimage.imread(image_filename, flatten=True)
# process mask to boolean
mask_image = ndimage.imread(mask_filename, flatten=True)
mask_image = mask_image[:, :] < mask_threshold
if not imaging.verify_mask(main_image, mask_image):
raise Exception("Mask does not fit image")
# automatic vector crop
crop_params = imaging.mask_crop_size(mask_image)
cropped_image = imaging.crop(crop_params, main_image)
cropped_mask = imaging.crop(crop_params, mask_image)
# make the dataset for this image
frame_data = thermal.thermal_image_to_dataset(
cropped_image, temp_range, pixel_range, id, cropped_mask,
(crop_params[0], crop_params[1]))
full_data.extend(frame_data)
i += 1
print("done!")
# save the dataset
print("Saving csv ...")
with open(csv_exportname, mode="w", newline="") as file:
write_obj = csv.writer(file)
write_obj.writerows(full_data)
print("Completed!")
if gui_enabled: easygui.textbox("Processing complete!")