It is an important topic in the field of medicine image to make research on Leukocyte microscope-image classification and recognition with computer pattern recognition technique. This thesis encloses the applied project about classification and recognition of leukocytes which are conventionally prepared with W-G stain. All studies focus on the key techniques of classification and recognition, such as segmentation of cells, feature extraction and the multi-class classification technique. The main contents are listed below:The first, original image is changed from RGB color space to HSI color space, and then the H-component image is processed by gray level transformations. Then images are segmented based on soft morphological watersheds. Its aim is to segment the nucleus, cytoplasm and background, even in the case of closely clustered cells. This procedure gets the borderline grads of the interesting areas.The second, a method describes image-region texture character based on multi-fractal is carried out. In addition, several features being able to reflect the leukocyte’s distinction at conformation, geometric, color and texture are extracted. Aimed to the problem that original feature is mass and redundancy in pattern recognition, a method of feature optimal based on genetic and simulated annealing algorithm is proposed.Finally, an application about HMMs is realized to solve the multi-class classification problem of leukocytes. To verify the performances of classifiers, the proposed method is compared with some normal classifiers. The result shows the method has a great perform on the field of classification and recognition of leukocyte microscope-images.

Related posts:

1. Studies on the Classification and Recognition of Red Blood Cell Microscope-Images Based on Morphology
2. Research into the Image Recognition Based on RBF Network
3. The Study of Clothing Image Emotion Classification Based on Multi-Features
4. Iris Recognition Method and Its Implementation Based on k-NN Classification Matching
5. Study on Object-oriented Remote Sensing Image Classification and Its Application