Ask Your Question

How does scale factor behave in HOG multiscale detector ?

asked 2015-06-24 09:09:12 -0500

erogol2 gravatar image

I try to experiment with HOG multi scale detector but cannot understand the exact behavior of the scale factor parameter. Could you explain what is the intuition behind this parameter?

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2015-06-24 11:43:52 -0500

Actually it does exactly the same as any other multiscale detection approach out there. It is used to create the image scale pyramid for multi scale detection using a single scale model.

It takes the following steps

  1. Start from a new test image, go with a sliding window approach (where window is model size) through the image and store all detections.
  2. Then downscale the image according to the scale factor, where 1.10 means you dowscale both dimensions (rows and cols) with 10%.
  3. Again perform detections.
  4. Warp the image detections back to the original size of the image using the downscale factor.
  5. Do this until one of the dimensions of the image reaches the largest model dimension (because then sliding window is no longer possible).
  6. Merge detections on similar locations over different scales.
edit flag offensive delete link more


@StevenPuttemans How is the merge done?

yogeshmurthy_ gravatar imageyogeshmurthy_ ( 2017-09-01 01:40:42 -0500 )edit

Start reading up on non-maxima-suppression :)

StevenPuttemans gravatar imageStevenPuttemans ( 2017-09-01 02:58:39 -0500 )edit

Using mean shift algorithm?

yogeshmurthy_ gravatar imageyogeshmurthy_ ( 2017-09-01 03:23:05 -0500 )edit

Start reading: IMHO this was just a simple google search ...

StevenPuttemans gravatar imageStevenPuttemans ( 2017-09-01 06:28:46 -0500 )edit
Login/Signup to Answer

Question Tools

1 follower


Asked: 2015-06-24 09:09:12 -0500

Seen: 715 times

Last updated: Jun 24 '15