Selection of window size in Adaptive thresholding

2
votes

I need to segment characters from following License plates (single row license plate and double row license plate) using cvAdaptiveThreshold(). Which is the best window size for doing adaptive threshold in first image and in second image. 

How is window size for adaptive threshold is determined ??

Image 1

Image 2

1
opencvimage-processingcomputer-vision

1 Answers

0
votes

Your example images are one of those kind on which researchers are trying to find robust algorithms. And in my opinion adaptive thresholding approach might not be the best option. The reason is as follows:-

In this technique the region within the window size is assumed to be uniform . Now in your case the characters size and distance between camera to license plate (Guessing the images to be of license plates) are important factor in determining correct window size. But those factors again are difficult to detect.

You might consider Local binary pattern based tecnique or look into this paper.

Yoon, Youngwoo, et al. "Best Combination of Binarization Methods for License Plate Character Segmentation." ETRI Journal 35.3 (2013).