python main.py register --camera_params "CameraParams/CameraParams.npz" --input_image "old_files/andrew photo video/reference messy.jpg" --output_image "output_script_test.jpg" --crop_method "none" --feature_method "SIFT" --model_output "model_script_test.npz"
Skipping image cropping as per the selected method.
Model saved to model_script_test.npz
python main.py detect --model_input "model_script_test.npz" --input_image "old_files/andrew photo video/second pic messy.jpg" --draw_match
Traceback (most recent call last):
File "~/ARCube/main.py", line 6, in <module>
parse_args_and_execute()
File "~/ARCube/parse_args.py", line 111, in parse_args_and_execute
detect(args)
File "~ARCube/parse_args.py", line 68, in detect
detector.detect_image(args.input_image, useFlann=args.use_flann, drawMatch=args.draw_match)
File "~/ARCube/detection.py", line 165, in detect_image
matches = self.matcher.knnMatch(des1, des2, 2)
cv2.error: OpenCV(4.10.0) /io/opencv/modules/core/src/batch_distance.cpp:282: error: (-215:Assertion failed) (type == CV_8U && dtype == CV_32S) || dtype == CV_32F in function 'batchDistance'
Got an error running the example script without flann, to reproduce: