Thanks for the amazing work. I am getting an error while running on a custom video. Its time duration is 6 seconds and I changed the parameters. but it shows me an error in the attached image.
Your help would be highly appreciated.

I've downloaded you project and changed the file_path attribute in the frame.py file, here's the code:

    file_path =  "Video/1min.mp4"# <== Upload appropriate file          
    video_out = "Output/1min.mov"

    frame =  FRAME( 
        ego_vehicle_offset = .15,
        yellow_lower = np.uint8([ 20, 50,   100]),
        yellow_upper = np.uint8([35, 255, 255]),
        white_lower = np.uint8([ 0, 200,   0]),
        white_upper = np.uint8([180, 255, 100]), 
        lum_factor = 118,
        max_gap_th = 0.45,
        YOLO_PERIOD = .25,
        lane_start=[0.35,0.75] , 
        verbose = 3)
    frame.process_video(file_path, 1,\
            video_out = video_out,pers_frame_time =144,\
            t0  =144 , t1 =150)#None)

Running the program, I got this error:

davidevitiello@iMac copilot % sudo python3 frame.py
2020-10-07 00:09:10.308022: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN)to use the following CPU instructions in performance-critical operations:  AVX2 FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2020-10-07 00:09:10.330339: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x7fc4fff30b10 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-10-07 00:09:10.330371: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Host, Default Version
Video/1min.mp4 WIDTH 640 HEIGHT 360 FPS 29.97 DUR 59.0 s
Traceback (most recent call last):
  File "frame.py", line 313, in <module>
    frame.process_video(file_path, 1,\
  File "frame.py", line 242, in process_video
    self.image = self.image[:frame_h_out,:,:]
TypeError: 'NoneType' object is not subscriptable

How I do get rid of it?

I do have Python 3.7.3 installed on my system an iMac

Dear Author,
When I run the last part of the code from frame import FRAME

file_path = "./videos/nice_road.mp4"# <== Upload appropriate file
video_out = "videos/output11.mov"
frame = FRAME(
ego_vehicle_offset = .0, # offset for self vehicle in frame
yellow_lower = np.uint8([ 20, 50, 110]), # LOWER YELLOW HLS THRESHOLD
yellow_upper = np.uint8([35, 255, 255]), # UPER YELLOW HLS THRESHOLD
white_lower = np.uint8([ 0, 140, 0]), # LOWER WHITE THRESHOLD
white_upper = np.uint8([180, 255, 100]), # UPPER WHITE THRESHOLD
lum_factor = 110, # NORMALIZING LUM FACTOR
max_gap_th = 0.45, # MAX GAP THRESHOLD
YOLO_PERIOD = 2, # YOLO PERIOD
lane_start=[0.2,0.5] , # LANE INITIATION
verbose = 3) # VERBOSITY
frame.process_video(file_path, 2,video_out = video_out,pers_frame_time =398, t0 =398 , t1 =698)#None) , I am getting the following error: "ZeroDivisionError Traceback (most recent call last)
in ()
14 lane_start=[0.2,0.5] , # LANE INITIATION
15 verbose = 3) # VERBOSITY
---> 16 frame.process_video(file_path, 2,video_out = video_out,pers_frame_time =398, t0 =398 , t1 =698)#None)

/content/drive/MyDrive/copilot/frame.py in process_video(self, file_path, fps_factor, video_out, pers_frame_time, t0, t1)
232 frame_h_out = int(frame_h*(1-self.ego_vehicle_offset))
233 print("{:s} WIDTH {:d} HEIGHT {:d} FPS {:.2f} DUR {:.1f} s".format(
--> 234 file_path,frame_w,frame_h,fps_actual,nb_frames//fps_actual
235 ))
236

ZeroDivisionError: float divmod()". Kindly help me out in the same.

Hi, Thanks for the wonderful work. I really appreciate it. I have a few doubts about the parameters used while running on video.

1. how do you describe lane start? is it the same coordinate to start detecting the lane in the bird view image?
2. which factors apart from frame's parameters do I need to change while running on the custom video?

Thanks!!

With coco dataset I can run the code. Now i want to use a custom dataset with just 3 classes?
What would be best approach to replace coco and integrate custom weights?

Any help would be appreciated- Thanks.

Hi, Thanks for the wonderful work. I really appreciate it. I have a few doubts about the parameters used while running on video.

1. how do you describe lane start? is it the same coordinate to start detecting the lane in the bird view image?
2. which factors apart from frame's parameters do I need to change while running on the custom video?

Thanks!!

I'm getting this error when I'm trying to run the code in colab where the LANE_DETECTION is called:
PERSPECTIVE TRANSFORMATION MATRIX COMPUTED

error Traceback (most recent call last)
in <cell line: 11>()
9 video_reader.set(1,pers_frame)
10 ret, image = video_reader.read()
---> 11 ld = LANE_DETECTION( image,fps,
12 yellow_lower = np.uint8([ 20, 50, 110]),
13 yellow_upper = np.uint8([35, 255, 255]),

1 frames
/content/lane_detection.py in calc_perspective(self, lane_start)
431 cv2.line(img, (int(x2), 0), (int(x2), self.UNWARPED_SIZE[1]), (0, 0, 255), 3)
432
--> 433 cv2.circle(img_orig,tuple(self.vanishing_point),10, color=RED, thickness=5)
434 cv2.circle(img_orig,tuple(orig_points),10, color=GRAY, thickness=4)
435 cv2.imwrite(self.temp_dir+"vanishing_point.jpg",img_orig)

error: OpenCV(4.8.0) 👎 error: (-5:Bad argument) in function 'circle'

Overload resolution failed:

• Can't parse 'center'. Sequence item with index 0 has a wrong type
• Can't parse 'center'. Sequence item with index 0 has a wrong type

Where I think the error is due to it not recognizing the tuple(self.vanishing_point) and thus giving the error.
Would like to know the fix to this

Hello, buddy, this project is really good, I am very interested in your research, I have three questions to ask:
1, I would like to ask this program can run under what operating system, can run on the PC?
2. What toolkits do you need to install before you can run the system?
3. Can you tell me the setup steps and the running command line?