A_particle_filter_based_fusion_framework_for_video

运用最小二乘方法的matlab影像匹配，经典算法(MATLAB image matching using least squares method, classical algorithm.)

极坐标和平面坐标的转换，在图像处理中可以很好解决在一个坐标下不能解决的问题。(Polar coordinates and plane coordinates conversion, in image processing can be solved in a coordinate system can not solve the problem.)

能量熵的计算，FMCW调频连续波雷达的测距测角，完整的图像处理课设，包含所有源代码，汽车图像。( Energy entropy calculation, FMCW frequency modulated continuous wave radar range and angular measurements, Complete class-based image processing, contains all of the source code, auto image.)

实现六自由度运动学逆解算法，汽车课设货车Matlab驱动力图程序，对信号进行频谱分析及滤波。( Six degrees of freedom to achieve inverse kinematics algorithm, Car class-based truck driver trying to Matlab program, The signal spectral analysis and filtering.)

opencv c++将彩色图像转换为灰度图像进行边缘检测，(opencv c++ to convert color images to gray-scale image edge detection,)

说明：  利用小波进行边缘提取，比较各种小波核函数的能力，体会“数学显微镜”的妙用，例子：车牌定位。(using wavelet edge detection, comparison of different wavelet function of the nuclear capability and experience "math microscope," the magic examples : positioning plates.)

说明：  改进了Criminisi算法，采用P-laplace算子作为数据项进行优先级计算，并有计算PSNR的功能(Improved Criminisi algorithm, using P-laplace operator as data items priority basis, and have calculated the functions of PSNR)

二值化处理图像，自动拟合椭圆并得到长短轴的大小。(Automatically fit the ellipse and get the size of the long and short axes)

多尺度LBP，代码根据论文“Learning Multi-scale Block Local Binary Patterns for Face Recognition”编写的(this code is obtained from the paper named "Learning Multi-scale Block Local Binary Patterns for Face Recognition")

带高斯白噪声的Kalman滤波Matlab代码，每一步都有详细的注释，和kalman滤波的五个公式对应。并有高斯白噪声的产生函数，function G = CreateGauss(E,D,M,N) 产生均值为E,方差为D,MxN的高斯白噪声矩阵(The Matlab code of Kalman filter with Gaussian white noise.)