1
I am working on a software that processes very large images, between 10GB and 30GB. The images are too heavy to open at once. I would like to know if there is any way to use Opencv to open the image in parts, as if to divide it into ten parts, applying the filters one part at a time, and in the end joining everything in a processed image.
I know that this would require some form of Buffer. Can anyone help me? I don’t have code examples because it’s just a question and I don’t even know how to do it.
Opencv alone is unable to upload a larger 4GB image and if it could take 4GB of memory RAM from its O.S. https://github.com/opencv/opencv/pull/1438 https://github.com/introlab/find-Object/issues/66 The community usually uses another lib to upload large images or cut the images into small numbered files and then process them in sequence. http://www.imagemagick.org/Usage/crop/#crop_tile
– gorn
An alternative is the scikit.
– danieltakeshi