基于小波变换的激光数字图像抗噪性检测系统设计

Design of laser digital image anti-noise detection system based on wavelet transform

针对传统系统存在检测完成时间过长、去噪性能效果较差、成本消耗高的问题。设计出基于小波变换和CIP算子的激光数字图像抗噪性检测系统。通过对激光数字图像进行分析,提取激光数字图像特征,降低图像特征矢量的维数,所设计的检测系统可以分别进行激光数字图像离线训练和在线检测,具有一定的再学习功能,利用小波变换对激光数字图像进行图像去噪压缩,对去噪后的压缩图像采用CIP算子进行定位,将激光数字图像定位结果利用BP神经网络进行识别检测。测试结果表明,与传统系统相比,本文系统检测完成时间较短、去噪性能效果较好。

Aiming at the problems of long detection time,poor de-noising performance and high cost consumption in traditional systems.A laser digital image anti-noise detection system based on wavelet transform and CIP operator is designed.Through the analysis of laser digital image,the feature of laser digital image is extracted and the dimension of image feature vector is reduced.The detection system designed can train and detect laser digital image offline and online respectively.It has a certain re-learning function.Wavelet transform is used to denoise and compress laser digital image.CIP operator is used to locate the compressed image after denoising,and BP neural network is used to recognize and detect the location result of laser digital image.The test results show that compared with the traditional system,the designed system has shorter detection time,better denoising performance.