降质图像高密度斑点批量智能识别仿真

针对降质图像中高密度斑点识别速度慢、识别位置不精准问题,提出了一种基于频域滤波法和相交测定中心法的降质图像高密度斑点批量智能识别技术。首先使用小波域内非线性软门限选取方法对图像进行去噪处理,选取小波域内非线性软门限,根据门限剔除图像冗余干扰,并确保滤除斑噪的同时保证边缘信息安全,然后使用相交测定中心算法对去噪后图像进行梯度识别,获得高密度斑点的中心面积,最后通过斑点细来设定阈值,从而完成降质图像高密度斑点批量智能识别的目的。仿真证明,相对于传统方法,所提方法在降质图像高密度斑点的批量智能识别上效率更高,并且识别精准度较高。

Simulation of High Density Speckle Batch Intelligent Recognition for Degraded Image

For the problems about slow recognition speed and inaccurate location of high-density speckle in de-graded image,this paper puts forward an intelligent recognition method of massive high-density speckles in degraded image based on frequency domain filtering and intersection detection center method.Firstly,the nonlinear soft thresh-olding method in wavelet domain was used to remove the noise in image,and the nonlinear soft threshold in wavelet domain was selected to eliminate the redundant interference of image.Meanwhile,it was necessary to ensure the edge information while filtering the speckle noise.And then,the intersection measurement center algorithm was used to discriminate the gradients of image after noise reduction and thus to obtain the central area of high-density speckles.Finally,the threshold was set by the speckles,so that the intelligent recognition for massive high-density speckles in degraded image was achieved.Simulation results show that the proposed method is more efficient and accurate in large batches of intelligent recognition of high-density speckles in degraded image than traditional methods.