摘要
探讨RGB和HSL彩色空间转换技术,提出利用HSL空间中H分量的颜色直方图对几种胃癌细胞的颜色特征进行提取和表达,并结合切片图像的颜色矩等参数综合表达其颜色特征,还探讨了BP网络分类方法中的参数的调整技术,实验显示该方法能较好地将不同种类的胃癌细胞区分开,达到了应用要求的识别率。
It probes the conversion technology between RGB and HSL color spaces and proposes a way to express the color features of gastric cancer cell by H componment histogram in HSL color space.Then by combining the color moments of submitting images with other color properties it express the integrated features of an image＇s color more details.Experiment shows that the proposed method can distinguish different types of gastric cancer cells well by subtly adjusting the parameters of BP neural network and can gain the required recognition rate in application.
出处
《江西科学》
2010年第5期626-629,672,共5页
Jiangxi Science
