摘要
纹理分析是肝脏B超图像识别的重要方法,本文利用多进制小波变换提取脂肪肝和正常肝脏的纹理特征,结合空域特性对B超肝脏图像进行分类,采用自适应浮动搜索算法进行特征筛选,经实验测试,取得了较好的效果。
In order to provide the science basis for making a diagnosis of fatty liver and increase the veracity, medical ultrasonic image of fatty liver and the normal liver were analyzed by the way of the multi-wavelet transformation. In this way can we get the texture Characteristic, combine the spatial domain and adapt searching method to classify them and receive good effect.
出处
《河北软件职业技术学院学报》
2008年第3期59-61,共3页
Journal of Hebei Software Institute
关键词
多进制小波
脂肪肝
B超图像
multi-wavelet texture
fatty liver
medical ultrasonic image
