摘要
将小波变换与圆谐联合变换相关器相结合，应用功率谱相减技术进行多目标检测，可以从输出面上直接看出系统对真目标的旋转不变检测及识别假目标的能力。分析和量化了输入景物噪声对结果的影响。计算机模拟实验表明这种方法明显改善了普通圆谐联合变换相关器的性能，具有良好的相关输出性能和抗噪声能力。
A method of multitarget detection is presented. It combines circular harmonic joint transform correlator (CHJTC ) with wavelet transform, and introduces power spectrum subtraction technique. Using this method we can find out directly the rotation-invariant recognition for the real pattern and the discrimination performance relative to the false pattern. The effect on the result of noise in the input scene is analyzed and quantified. Computer simulation verifies that it can enhance obviously the performance of general CHJTC, and yield a better correlation output and higher capability of anti-background noise.
出处
《应用激光》
CSCD
北大核心
1999年第3期97-100,共4页
Applied Laser
关键词
光学图像识别
圆谐
相关器
小波变换
多目标检测
optical pattern recognition, circular harmonic joint transform correlator,wavelet transform, power spectrum, multitarget detection
