The processing of blood cell micro-flow image and objects recognition

A processing method of blood cell flow image in micro rectangular capillary is proposed and the recognition and computation procedures are also discussed. According to the characteristics of cell micro-flow image, the image is filtered and enhanced in frequency spectrum, and mathematical morphology operations are applied to recognize and segment the flowing objects in image. With the segmented results, the edges of cells in image are recognized and reconstructed to calculate their locomotors conditions. Then the flow parameters of every cell are computed, including position, area, perimeter, and orientation angle et al. The computed results reveals that the blood cell micro-flow image can be processed with background removing, contrast enhancing, and morphological operating, and the objects recognized match the original cell image perfectly. This processing method can also been used in other micro-flow phenomena researches.