array = np.array(image_gray)
# 1. gaussian blur to remove high frequency noise
array = Filter.gaussian_filter(array)
# Image.fromarray(np.uint8(array)*255, 'L').show()
# 2. apply sobel operator to find gradient
SobelX, SobelY = Filter.sobel_filter(array)
# Image.fromarray(np.uint8(SobelX)*255, 'L').show()
# 3. non-max supprssion
suppressed = Detector.non_max_suppression((SobelX, SobelY))
# Image.fromarray(np.uint8(suppressed), 'L').show()
# 4. canny detection
canny = Detector.canny_detector(np.uint8(suppressed) * 255)
canny_image = Image.fromarray(np.uint8(canny) * 255, 'L').convert("RGB")
# Image.fromarray(np.uint8(canny) * 255, 'L').show()
# 5. Hough transformation
vertical_lines = Detector.hough_lines(canny, 180, 452)
horizontal_lines = Detector.hough_lines(canny, 90, 193)
# output image
Detector.drawLines(canny_image, vertical_lines[0], vertical_lines[1])
Detector.drawLines(canny_image, horizontal_lines[0], horizontal_lines[1])
canny_image.save(args.output_im)
canny_gray = ImageOps.grayscale(canny_image)
canny_gray.save(args.output_im)
##canny_img = np.uint8(canny_image)#canny_image.load()
