Multiple bounding boxes around Object (OpenCV)

Asked Aug 09 '22 at 11:39

Active Aug 09 '22 at 15:12

Viewed 440 times

I am new to ML, and OpenCV and was trying to perform object detection of a car using the pre-trained model on Coco Dataset, DNN. I am receiving multiple bounding boxes around the car.

The code and image are attached for reference.

cap = cv2.imread(r'D:\car.jpg')

while True:
    (classIds, scores, bboxes) = model.detect(cap)
    for classId, score, bbox in zip(classIds, scores, bboxes):
        (x, y, w, h) = bbox
        className = classes[classId]

        if className == 'car':
            cv2.putText(cap, className, (x, y-10), cv2.FONT_HERSHEY_PLAIN, 2, (200,0,50), 2)
            cv2.rectangle(cap, (x, y), (x+w, y+h), (33, 191, 130),3)

    cv2.imshow('abc', cap)
    if cv2.waitKey(33) == 27:
        break

edited Aug 09 '22 at 15:12

Christoph Rackwitz

11,317
4
27
36

asked Aug 09 '22 at 11:39

Ayush Porwal

3

Look for "Non Maximal Suppression". It removes overlapping boxes – Jeru Luke Aug 09 '22 at 11:52
What is the purpose of the `while True` loop? – Markus Aug 09 '22 at 12:10
@Markus I couldn't perform the exit without that. – Ayush Porwal Aug 09 '22 at 12:21
Non-Maxi**mum** Suppression but google will correct that – Christoph Rackwitz Aug 09 '22 at 15:13
@Ayush Porwal. What if you move outside of the if/else condition in line 11? – toyota Supra Aug 09 '22 at 17:55
@toyotaSupra I did that but it's not helping – Ayush Porwal Aug 10 '22 at 06:38
Have you tried ready to use "non maximum suppression"-functions from the net as Jeru Luke suggests, like [here](https://pyimagesearch.com/2015/02/16/faster-non-maximum-suppression-python)? Do you still have issues? – Markus Aug 10 '22 at 08:01
@Markus I did try NMS. It helped thanks. The accuracy is low though. Can you please suggest something regarding the same? TIA – Ayush Porwal Aug 10 '22 at 08:39
@Ayush Porwal: What do you mean by low accuracy? – Markus Aug 10 '22 at 08:50
@Markus If there are N cars in an image, around N/2 are detected. – Ayush Porwal Aug 10 '22 at 08:59
@Ayush Porwal. What if you move outside of the for looping in line 11? – toyota Supra Aug 10 '22 at 09:04
1

@toyotaSupra yes. I did actually after you informed it. The boundary box issue is resolved with the help of NMS. But I am unable to detect with good accuracy. – Ayush Porwal Aug 10 '22 at 09:12
Can you post second image? – toyota Supra Aug 10 '22 at 09:14
@toyotaSupra Markdown ![sample image](https://ibb.co/QQtNh66) – Ayush Porwal Aug 10 '22 at 09:25
Notice. Using the same video that you posted previously, But no other videos.. – toyota Supra Aug 10 '22 at 09:31
@toyotaSupra Almost no other images. Is it because of the dimensions of the image? – Ayush Porwal Aug 10 '22 at 09:35

Multiple bounding boxes around Object (OpenCV)

0 Answers0