摘要
目前分形理论已被广泛应用于医学图像识别.本文利用了改进的差分盒计数法对肝脏CT图像进行了分形维数的分析,表征了肝脏图像的纹理特征,则可识别医学图像肿瘤信息.医学图像识别的结果可以直观的方式显示给临床医生,从而辅助医生提高医疗诊断的准确性和科学性.
At present,fractal theory is widely applied in medical image recognition.In this paper,by using the differential box counting we analyze the fractal dimension of liver CT images,depicts the texture characteristics of the medical image,so medical image tumors information can be recognized.The medical image recognition results will be shown to the clinician in intuitive way,thus the accuracy and scientificalness of medical diagnosis will be improved.
出处
《南华大学学报（自然科学版）》
2010年第1期56-58,共3页
Journal of University of South China：Science and Technology
基金
衡阳市科技计划基金资助项目(2008KJ010)
关键词
图像识别
分形维数
纹理特征
image recognition
fractal dimension
texture feature
