摘要： 针对骨髓细胞图像的特点，采用数学形态学的方法对图像进行了处理，获得了不同类型细胞核的准确边缘。对于获得的边缘图像，采用两级神经网络，利用基于神经网络的PCA 算法获得图象的3 个主分量，然后采用模拟退火算法和BP 算法进行细胞的分类识别，获得了较好的识别效果。
关键词:
数学形态学；主分量分析；模拟退火算法；BP 网络
Abstract: Aiming at the characteristics of the marrow cell image, the paper adopts morphology methods to process image so that accurate edgesof different kinds of cells can be detected. This paper adopts two neural networks, using principal component analysis(PCA) based on neural networkto acquire three principal components and using simulated annealing(SA) and BP network to class and recognize the marrow cells.
Key words:
Morphology; Principal component analysis (PCA); Simulated annealing (SA); BP Network
侯振杰，麻硕士，裴喜春，潘 新. 基于形态学与神经网络的骨髓细胞识别研究[J]. 计算机工程, 2006, 32(11): 203-205.
HOU Zhenjie，MA Shuoshi，PEI Xichun，PAN Xin. Study on Recognition of Marrow Cell Image Based on Morphology and Neural Networks[J]. Computer Engineering, 2006, 32(11): 203-205.