利用小波变换和边缘限制条件进行水域分割

阐述了一种利用小波变换和边缘限制条件进行水域提取的新算法，它的特点是先用小波变换将影像变换到不同尺度层上；然后在不同尺度层上统计影像的特征值，以形成影像的特征数据；接着利用Facet边缘检测算法在特征数据上检测特征边缘存在的情况，从而判断像素是否属于水域，与传统的水域提取方法相比，它用到了不同频率上水域灰度信息，从而更准确地刻画了水域的灰度特性，试验结果表明，利用小波变换和边缘限制条件对水域分割效果较好。

Application of Segmentation Algorithm Based on Wavelet Transform and Limiting Condition of Margin in Water Area Extraction

Aiming at the randomicity and complexity of residential area extraction from black and white aero photo, and to improve the accuracy of boundary locations and region homogeneity as well as to reduce the error rate in water area segmentation, this paper points out a new algorithms that extracting the water area in wavelet transform and limiting condition of margin. This new technique's characteristic is to transform the initial image to pyramid structured in wavelet transform firstly, then calculate eigenvalues from the different scale, and forming the features of aero photo. After that we use the method of Facet margin detection to detect if there is the boundary, then to decide what pixels belong to. All technical points are clearly described and presented in detail. Compare with the typical traditional method, the present approach use different scale's gray information of water area, so it can describe the features of water area more accurately. Experiences show that the present method of wavelet transform and limiting condition of margin shows visible improvements both in diminishing segmentation error, and in increasing boundary precision and region harmony.