Haar Cascade vs Hog Detection

14,829

Solution 1

HOG is usually better for human detection, than Haar. I have only experience in this so I thought I'd give some input on that. However, the limitation of HOG is that the human must be within a "perfect" area on the screen. Too close, it won't detect the human. Too far, it won't detect the human.

I have had better luck with HOG than Haar. Haar gave me too many false positives.

Solution 2

I have been trying to use HAAR to detect human, and it turns out to give too many false positives. I think HAAR is only suitable for face or eye detection.

since your camera is in the sky, the human is pretty small in the image and got a whole body shape. HOG would be a better choice.

Solution 3

You need to change scale factor and minimum neighbours in HAAR cascade which is not same for all the image. So it's better to use HOG.

Share:
14,829

Related videos on Youtube

user2039318
Author by

user2039318

Updated on September 09, 2022

Comments

  • user2039318
    user2039318 about 1 year

    I have been working around with OpenCV for few days now and I have a project where I should detect cars and humans from the sky.

    So here are my inputs:

    • A moving camera in the sky (embedded on a quadcopter) which is gonna capture frames.
    • A set of objects I should detect (humans and cars)

    And here are my output:

    • A detection of those objects outlined by a rectangle or some contours

    Based on that, my question is as follows: Which one between Haar Cascade and Hog Detection would you recommend to do so and why? Or any else?

    Many thanks for your answers

    • LovaBill
      LovaBill over 10 years
      In case your camera faces downwards, none. Upright position with frontal view was used only in the pre-trained models. You'll have to train your own model to improve the results. In that case, test both.
  • saurabheights
    saurabheights almost 8 years
    False positives can be discarded using skin detection(color-based) if we are dealing with color images.