Application of artificial immune algorithm in image segmentation based on immune field

Image segmentation is one of the classic problems in image processing and computer vision field. Existed algorithms do not always reach a satisfactory purpose in fuzzy image segmentation. This paper is inspired by new development of medical immunology and proposes an artificial immune algorithm based on immune field. First, the article gives the concept of innate immune field, adaptive field and the immune field by learning from the operating mechanism between innate immune system and adaptive immune system. Second, the paper builds an artificial immune network of combination between innate immune and adaptive immune to divide the antigen feature space. This novel artificial immune algorithm is used for segmenting of object, background and thermal diffusion region in sheltered infrared image. Experimental results show that the method we proposed can solve the problem of incomplete target and edge distortion, and have a comparatively satisfying result with comparison to some segmentation methods, such as immune template, Prewitt operator, Sobel operator and negative selection.