CHAOS_PSO_TEF

图形处理的部分，能进行图像的中值过滤处理(To the median filter for image processing)

说明：  给定两张图片，利用ssim来计算两张图片他们之间的ssim相似度(compute ssim between two images)

说明：  Z. Liang, J. Xu, D. Zhang, Z. Cao, L. Zhang, "A Hybrid L1-L0 Layer Decomposition Model for Tone Mapping," in CVPR 2018.

建立了混合高斯模型，可以对静止背景下运动目标进行检测。(Gaussian mixture model is established, you can still detect moving target in the background.)

一种可适用于复杂光照环境的低照度图像增强算法，即基于内容感知的自适应对比度提高的增强算法。引入了改进的伽马矫正，可以根据低照度图像的明暗程度自适应的对图像进行预处理。在实验过程中还发现，虽然改进算法有效避免引入大量噪声，但仍有少量噪声存在，为了达到更好的视觉效果，利用改进的双边滤波对增强的图像进行处理，获得最后的增强图像。实验效果显示，该算法可对低照度图像实现理想的增强效果。(A can be applied to complex light environment of low illumination image enhancement algorithm, which based on the content of perception adaptive contrast enhancement algorithm. Introduced the improvement of gamma correction, can according to the low illumination image of light and shade degree adaptive to image preprocessing. In the process of experiment also found that although the improved algorithm effectively avoid introducing a lot of noise, but there still have a small amount of noise, in order to achieve better visual effect, use of improved bilateral filtering to enhance the image processing, finally enhance image. The experimental results showed that the algorithm can achieve a low illumination image enhancement effect. )

基于低频融合算法和小波变换编写的图像融合Matlab程序，包含了基本的小波图像分解与重构程序。(Fusion algorithm based on the low-frequency and wavelet transform image fusion prepared Matlab procedures, including the basic wavelet image decomposition and reconstruction process.)

说明：  无人机影像ORB特征提取和RANSAC算法剔除(UAV image ORB feature extraction and RANSAC algorithm elimination)

这是一个印度人编的程序，用于两图像单应矩阵和变换结果的求取(This function estimates 2D-2D plane projective homography between two perspective images using Direct Linear Transformation, RANSAC and Levenberg Marquardt optimisation.)

用来计算水印算法中的psnr值，非常好用一直在用这个(Watermarking algorithm used to calculate the value of psnr, very easy to use )

SAR成像处理的过程是从回波数据中提取目标区域散射系数的二维分布，本质上是一个二维相关处理过程,因此最直接的处理方法是对回波进行二维匹配滤波，但其运算量很大，再加上SAR的数据率本来就高，这使得实时处难于实现。通常，可以把二维过程分解成距离向和方位向两个一维过程，Range-Dopper Algorithm（简称RD算法）(SAR imaging process is two-dimensional distribution of the scattering coefficient of the target area extracted from the echo data, in essence, is a two-dimensional process, the most direct approach is two-dimensional matched filtering of echo, its computationalthe great plus on SAR data rate was high, which makes real-time processing is difficult to achieve. Typically, two-dimensional process can be broken down into the range and azimuth to a two-dimensional process, the Range-Dopper Algorithm. (Referred to as the RD algorithm))