利用小波变换局部模极大值进行实时目标定位的应用

提出了一种基于小波变换局部模极大值的边缘检测方法．利用B样条小波构造出一个边缘检测算子，采用多孔算法进行边缘提取，减少了图像识别与测量中必要信息的丢失，并且提出了自适应阈值的方法，较好的解决了图像边缘的提取精度与对噪声抑制能力之间的矛盾．实验中对比了几种传统边缘检测算子和小波模极大值方法的在边缘检测效果、抗噪性和运算时间等方面的性能，本算法对加噪图像和模糊图像都获得比传统算法更好的检测效果，证明了基于三次B样条小波的局部模极大值方法检测目标边缘的有效性和正确性．

Application of Real-Time Object Orientation Based on Local Modulus Maxima of Wavelet

In this paper,a method to detect image edges based on local modulus maxima of wavelet is proposed.An edge detection operator is constructed by B-spline wavelet.Algorithm a'trous is adopted for image edge detection to prevent the loss of meaningful image information while recognizing and measuring.The contradiction between the detection precision of image edge and restraining ability on image noise is well resolved by applying a method of adaptive threshold.By simulation,several traditional edge detection operators are compared with the method of local modulus maxima on wavelet on the detection effect,anti-noise capacity and running time.The edge detection effect for Gauss noise-added image and blurred image is better than traditional operators.Experiments prove the veracity and effectiveness of this method with cubic B-spline wavelet.