Image Segmentation Based On Grayscale Morphological Reconstruction

With the development of technology of the computer and digital image technology, image method has already become a kind of important means of measuring particles, the purpose is to obtain particle appearance message of outline accurately from particle image. In the particle image segmentation, the objects are always touched or overlapped into each other so that it is necessary to separate or split them into single ones.Compared to other image processing methods, the mathematical morphology method has the unique advantage in image processing, in this paper, the watershed algorithm of mathematical morphology is used for image segmentation.This paper suggests an improved marker-based watershed image-segmentation method to reduce the over-segmentation of the watershed algorithm. This algorithm could segment more effectiv ely the touched kernels and inhibit the over-segmentation phenomenon.The main contributions from this paper include :1. A new algorithm of image segmentation is introduced, which is based on grayscale morphology reconstruction and effectively overcome th ese disadvantages of only using thresholding segmentation. This method improves the computation efficiency and attains higher precision.2. A marker-based watershed image-segmentation method is proposed. Mark the gradient image of soybean in the full use o f soybean's image information. Then the watershed algorithm is applied to the modified gradients by the markers, can more efficient reduce the over-segmentation ratio of the watershed algorithm. The experimental results show that the improved watershed algorithm can obtain better segmentation algorithm than simply the watershed for image segmentation applications, it has a good prospect applying to image processing.