Research On Cell Image Segmentation Based On Mathematical Morphology

Mathematical morphology offers a unified and powerful approach to numerous image processing problems by its precise theory system and intuitionistic geometry. Recently, images segmentation based on mathematical morphology has mainly focused on two parts: one is edge detection based on mathematical morphology; the other is region segmentation based on watershed transform. This paper presents a cell image edge detection method based on morphological reconstruction algorithm and an improved morphological reconstruction and marker-based watershed image segmentation method. The result is satisfactory when the two methods is applied to cell image processing. The main research work of this paper is as follows:1. The background and significance of the study on image segmentation are explained, and the status of the cell image analysis system at home and abroad is introduced. The concept of image segmentation and the characteristics of cells image as well as the existing segmentation methods of analysis are reviewed. This paper also presents the basic theory of morphology; the principle and the application of morphological reconstruction.2. It was found through the experiment that the simple use of morphological opening and closing operation will lead to edge information discontinuous and affect the accuracy of segmentation. So the morphological reconstruction filter has been designed to solve the conflict of predigesting image and maintaining the edge information. It is widely used and satisfactorily resolved the issue.3. Morphological opening and closing reconstruction filter operation is directed to poor filter effect of serious noise image. The method is improved via putting forward a morphology alternating sequential filter by reconstruction. It is so satisfied that it applied to cell image filter. 4. Key to select structure of mathematical morphology. Furthermore if the structural element is too small it can not own good anti-noise performance, and if the structural element is too large, it would destroy the figure of cells either. A self-adaptive method based on multi-structure and iterative dilatation combination is presented. These advantages have been adopted; suggest a multi-structure and multi-grade combination edge detection algorithm based on morphology reconstruction. The experimental results show that the new algorithm is better than the classical edge detection and classical morphological edge detection algorithms. Compare to the method of reference[59] supported, we adopt some more often used edge detecting operators to detect the cell's edge to obtain better result than the method of just using area morphology to detect cell image edge.5. I delve into the watershed algorithm, suggest a new marker-based watershed algorithm for cell image segmentation based on mathematical morphology, which can be used to solve the traditional over-segmentation. With the original method of simulation experiments, the method can obtain more accurate segmentation results. Compared with other tradition methods, this system requires fewer computations and simpler parameters and can more efficiently reduce the over-segmentation ratio of the watershed algorithm.6. Due to requirement of research work, I rewrote function operator expression of traditional mathematical morphology theory.7. In this paper, a new algorithm is presented to solve two key problems of cell image segmentation. The ways to improve the auto-adaptive capability of the structural elements and to transform 2D cell image to 3D reconstruction can be used to obtain more valuable medical knowledge. These will be selected as the direction of the further study.