I use template matching to detect a specific pattern in image.The shift determined is very shaky. Currently I apply it to R,G,B channel separately and average the result to obtain float values.Please, suggest how to obtain subpixel accuracy. I was planning to resize the image and then return the data in original scale, please suggest any other better method
$\begingroup$
$\endgroup$
3
-
2$\begingroup$ Are you looking to measure details smaller than a pixel by looking at several different instances of the pattern that have different offsets from the pixel grid? Or are you looking to use the arrangement of the red, green and blue subpixels to improve accuracy with a single instance of the pattern? $\endgroup$– trichoplax is on Codidact nowCommented Mar 3, 2016 at 14:48
-
$\begingroup$ I'm trying to measure details smaller than a pixel so by several instance you mean shifting it to left or right and then match .Thanks for your response. $\endgroup$– Nikita ChopraCommented Mar 4, 2016 at 3:24
-
$\begingroup$ By several instances I mean an image with a repeat pattern as described here. By red, green and blue subpixels I mean taking advantage of the placement of different colours within a single pixel (the pixel geometry) as described here. It sounds like you probably mean the first approach. $\endgroup$– trichoplax is on Codidact nowCommented Mar 4, 2016 at 12:30
Add a comment
|