Will splitting an image into small blocks for processing help in extraction of edges?
I split an image into several blocks to further process it.
Now I am hardly noticing any difference while applying filters to this small block compared to applying filter to entire image to extract edges.
Now I'm confused, in what type of situations do splitting of images into blocks actual help?
No.
"Now I am hardly noticing any difference while applying filters" Which differences ?
@LBerger I hardly notice any difference after applying the filter like for edge extraction(blur image and then extract edge).
Have you got an example with numerical values ? Of course there is some difference near image border
What do you expect from splitting the image into blocks?
If you try to speed-up processing, blocks will not speed-up processing in most cases. For image filtering you can try to split the image in stripes instead. The Canny algorithm is already parallelized, depending on OpenCV version (see here and scroll down a bit) and can't be parallelized from outside. This results in wrong block/stripe border handling.
If you want to filter the image bocks isolated you should pass the flag BORDER_ISOLATED to the filter function
@matman My AIM is to achieve a better edge extraction from the objects in the image.
"Now I am hardly noticing any difference while applying filters" still waiting for numerical values...
and which filter do you use?
@LBerger What do you mean by numerical values? Median Blur, then Gaussian blur and then Canny, in this specific order.
Only differencewhen you divide image in small block it is the border. that's all and it does not help at all to process edges. If you think so write a paper