GRAY LEVEL THRESHOLDING IN BADLY ILLUMINATED IMAGES

Most gray level thresholding methods produces very good results in situations where the illumination gradient in the original raster image is regular and not too large. In other cases, such as a large linear change in illumination, a satisfactory bilevel image cannot be produced. One approach is to locate object pixels first, and then grow regions around these, generating a threshold value for each pixel. It can be assumed with a high degree of assurance that pixels at locations having a high gray level gradient form a part of an object, and that nearby pixels with similar gray levels will also be part of the same object. If regions are grown carefully, a threshold image will result.