Relative Technology And Realization Of DNA Microarray Lmages Recognition

Image- recognition is the advanced step of medical image processing. It is widely used in many research fields. Image segmentation and pattern recognition are the key technologies of image recognition.In this paper, image recognition is deeply researched. At first, ISODATA dynamic clustering is imported to median filtering algorithm. The algorithm can eliminate serious yawp noise and retain image detail.Then in this paper, edge detection and image segmentation are deeply researched. A multi-threshold image segmentation method based on C-means clustering is proposed. A image segmentation method based on Fuzzy C-means clustering is also proposed.The advantage and inferior position of image recognition between based on fuzzy theory and based on BP Algorithm are researched. Then some improved Algorithms and mixed algorithms are proposed. In addition, a Genetic Algorithms is proposed to detect geometric graphics in an image. In contrast with traditional Hough Transform, the algorithm uses less time and smaller space.At last, a software system used in DNA microarray images recognition is formed by C++ Builder in this paper. The system can be well used in many other practical fields of image processing and image recognition.