Erythrocyte Recognition System Design And Realization

Blood cell counting is an important item for patient's blood routine examination. The result of counting the white blood and red blood gives important information about patient's health status and plays an important role in diagnosis. At present, electric resistances and lights spread methods are widely adopted in automatic counting systems. Under the two above situations where the human vision can't arrive, we can only perform regular classification.Although the technology of the red cell examination has been developed, usually the level of the perfect standard is not included. And current red cell examination software generally doesn't include the automatic statistic function and the recognition function such as the"blood corpuscle image character bulletin system". The function of cell classification and statistics for current cell examination systems are still depends on the traditional manual ways.Aim to solve the above problems, this thesis designed and implemented a red cell automatic processing system. The system can perform the cell sampling, call segmentation and cell automatic statistics and analysis. The main functions of the system are: utilizing and gathering the erythrocyte picture obtained from image acquisition system, performing the pattern processes such as cutting, hole packing, corroding and so on, and finding out the central point to count the cell quantity. The system works well and improves the efficiency. The experiment test which uses different erythrocyte pictures indicates that the discrimination rate is about 90%. The algorithm proposed in the thesis can achieve high recognition rate and does important auxiliary diagnosis for morphology of red cells.