amartya-k / vision

This is a simple vehile detection project using Computer Vision

vision

This is a simple vehile detection project using Computer Vision

Here I used background subtractions methods of OpenCV Library of Python and some morphological transformation for accuracy. But this is for only static cameras.

At the corner of the video you can see the count of the vehicles which gets recorded when they cross a predefined limit. For the calculation of the object coordinates and object ids I defined a class called vehicles.py

For running it, just download all files and run the main.py Thank you.

vision's Issues

How do I save the output video?

I want to save the output video with detection results.
How should I add the command line for saving?

Error while running the file

Traceback (most recent call last):
File "C:/Users/mahe/Desktop/vision-master/main.py", line 80, in
mask=cv2.morphologyEx(mask,cv2.MORPH_CLOSE,kernalC1)
TypeError: Expected cv::UMat for argument 'kernel'

I didn't what is the error.Could you please he,lp me with the issue

i faced some error

While executing the main I got the following error, please help me out with this

File "main.py", line 80, in <module>
    mask=cv2.morphologyEx(mask,cv2.MORPH_CLOSE,kernalCl)
TypeError: Expected Ptr<cv::UMat> for argument '%s'

ValueError: not enough values to unpack (expected 3, got 2)

I get an error then running main.py

(qrcode) y700@jekson:~/projects/cardetect/car/vision$ python main.py
Red line y: 288
Blue line y: 192
Traceback (most recent call last):
  File "main.py", line 85, in <module>
    _, countours0,hierarchy=cv2.findContours(mask,cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_NONE)
ValueError: not enough values to unpack (expected 3, got 2)