The present invention provides a unique method, apparatus, system and storage medium for image binarization process, wherein an image to be processed is divided into a plurality of sub-images and an binarization threshold for each of the sub-images is determined based on the gray-levels of the edge pixels detected in respective sub-image. The image processing method of the present invention comprises: calculating a gray-level statistical distribution of pixels of an image, detecting edge pixels in an image based on an edge criterion corresponding to the gray-level statistical distribution, dividing the image into a plurality of sub-images; determining a binarization threshold for each of the sub-images based on the gray-levels of edge pixels detected in the sub-image; and performing binarization process for each of the sub-images based on the binarization threshold of the sub-image.