Cal_entropy

图像去雾代码，参考文献《An Investigation of Dehazing Effects on Image and Video Coding》(matlab code for dehazing，reference：<An Investigation of Dehazing Effects on Image and Video Coding>)

说明：  对不同噪声强度的加噪图像，用粒子群优化算法优化结构元素，找到最优结构元素，对图像进行去噪操作，使恢复的图像达到最大峰值信噪比(Noise intensity for different images with noise, using particle swarm optimization algorithm to optimize the structure elements, find the optimal structure element, the image de-noising operation to restore the image to a peak signal to noise ratio)

泊松融合在某种意义上解决了图像融合的问题，但其在融合的过程中仅仅考虑了融入图像的梯度，而没有利用背景图像的梯度，因此，在使用泊松融合的过程中，需要对融入区域的标记较为准确，否则，会造成背景纹理的丢失，视觉上可以明显觉察图像融入的痕迹。因此，对公式稍作修改，在考虑前景的同时考虑背景的变化，可以得到更优的结果。(Poisson convergence in a sense solved the problem of image fusion, but it only considers the integration of image gradient in the integration process, but not the use of a background image gradient, therefore, the use of the Poisson integration process, the need for more accurate marker into the area, otherwise, it will result in the loss of background texture, image integration can clearly perceive the visual traces. Therefore, the formula slightly modified, in considering the prospect of taking into account the changes in the background, you can get better results.)

混合高斯模型使用K（基本为3到5个） 个高斯模型来表征图像中各个像素点的特征,在新一帧图像获得后更新混合高斯模型,用当前图像中的每个像素点与混合高斯模型匹配,如果成功则判定该点为背景点, 否则为前景点。通观整个高斯模型，他主要是有方差和均值两个参数决定，,对均值和方差的学习,采取不同的学习机制,将直接影响到模型的稳定性、精确性和收敛性。由于我们是对运动目标的背景提取建模，因此需要对高斯模型中方差和均值两个参数实时更新。为提高模型的学习能力,改进方法对均值和方差的更新采用不同的学习率 为提高在繁忙的场景下,大而慢的运动目标的检测效果,引入权值均值的概念,建立背景图像并实时更新,然后结合权值、权值均值和背景图像对像素点进行前景和背景的分类。(Gaussian mixture model using K (essentially 3-5) Gaussian model to characterize the features of each pixel in the image, in the image of the new frame for updated Gaussian mixture model, with each pixel in the image with a Gaussian mixture current model matching, if successful, determined that the point of the background points, otherwise the former attraction. Throughout the entire Gaussian model, he mainly has two parameters determine the variance and the mean, the mean and variance of the study, to take a different learning mechanism, will directly affect the stability, accuracy and convergence model. Since we are moving object extraction of the background modeling, so the need for the Gaussian model variance and mean two parameters real-time updates. In order to improve the learning ability of the model, an improved method for updating the mean and variance of different learning rates to improve in the busy scene, large and slow moving object detection results, the introduction of)

本程序包是关于图形处理和人脸识别的,包括人脸Gabor特征提取，canny算子，水线阈值方法等．大家可以一起参考(The package is on the graphics processing and face recognition, including face Gabor feature extraction, canny operator, waterline threshold methods. Together we can make reference)

实现SURF特征点描述子算法，并使用Oxford Dataset图像自动进行精度评价(SURF feature descriptor and matching precision evaluation)

说明：  文件说明： a1:将二进制文件转为Matalb数据 并批量保存为mat格式 a2:mat数据批量转为图片(RGB三通道） a3:RGB图片批量转灰度图、批量改变图片大小 a4:动图制作 a5:用训练好的网络对图像语义分割(Document description: A1: Convert binary files to Matalb data and save them in batch as mat format A2: mat data batch to picture (RGB three channel) A3:RGB Picture Batch Grayscale, Batch Change Picture Size A4: Motion Mapping A5: Segmentation of image semantics using trained networks)

应用条件随机场进行图像分割 使用方法：将pydensecrf-py27或pydensecrf-py35重命名为pydensecrf，放入C:Program FilesAnaconda3Libsite-packages 运行：python inference.py im1.png anno1.png out.png(Renameing the file "pydensecrf-py27" or "pydensecrf-py35" to the new name "pydensecrf" Puting this file "pydensecrf" in this directory "C:Program FilesAnaconda3Libsite-packages" cd examples/ python inference.py im1.png anno1.png out.png)

FMCW调频连续波雷达的测距测角，能量熵的计算，ICA（主分量分析）算法和程序。( FMCW frequency modulated continuous wave radar range and angular measurements, Energy entropy calculation, ICA (Principal Component Analysis) algorithm and procedures.)

说明：  多目标跟踪，动态背景下实现多目标跟踪算法，算法优良，实时性和鲁棒性都比较好(Tracking, dynamic context to achieve multi-target tracking algorithm, algorithm quality, real-time and better robustness than)