rubik

image thesis introduction

edfa edfrnrt utnuh u h u yhu ub(unènn unu(nu(

matlab实现的RGB空间转换到HSI空间，以供参考(matlab realize RGB space conversion to HSI space for reference)

在毕设过程中，网上搜集来的基于matlab多点约束的波束形成算法，自己做了改进和注释(capon beamforming，multipoint restriction)

使用mtcnn批量检测人脸特征点，检测五个人脸关键点（眼睛中心、鼻尖、嘴角）(Batch detection of face feature points using mtcnn)

说明：  蒙特卡洛模拟光子在组织中的运动 matlab，内有详细的蒙特卡洛模拟光子在组织中的运动内有详细的蒙特卡洛模拟光子在组织中的运动(Monte Carlo simulation of the movement of photons in the organization MATLAB, there are detailed Monte Carlo simulation of the movement of photons in the organization there are detailed Monte Carlo simulation of the movement of photons in the organization)

说明：  用matlab将hanning窗、hamming窗、Blackman窗画在同一个图中。语音分帧，加窗，短时能量，短时过零率，端点检测(Using matlab to hanning window、hamming window and Blackman window draw in the same graph.Voice sub-frame, add windows, short-term energy, short-term zero-crossing rate, endpoint detection)

提取信号七个基于瞬时信息的特征：零中心归一化瞬时幅度功率谱密度的最大值,零中心归一化瞬时幅度绝对值的标准偏差，零中心非弱信号段瞬时相位非线性分量绝对值标准偏差，零中心非弱信号段瞬时相位非线性分量标准偏差，零中心归一化的非弱信号段瞬时频率绝对值的标准偏差，一个信号段的归一化瞬时频率功率谱密度的最大值，根据信号 QPSK 和16QAM在 XI 轴投影的不同表现，提出特征参数。(Extraction based on instantaneous information signal of seven characteristics: zero center normalized instantaneous amplitude of the maximum power spectral density, zero-center normalized standard deviation of the instantaneous rate of absolute zero center section of non-weak signal instantaneous phase nonlinear component absolute value of standard deviation, non-zero center section of the weak signal non-linear component of the standard deviation of the instantaneous phase, zero-center non-normalized absolute value of the weak signal section of the standard deviation of the instantaneous frequency of a signal segment of the normalized power spectral density of instantaneous frequency maximum, according to the signal QPSK and 16QAM in the XI-axis projection of the different performance of the proposed characteristic parameters.)

该文件为matlab环境下的批处理程序，可以处理同一个文件夹下面的相同类型的文件(The file is a batch program matlab environment that can handle the same folder following the same type of file)

Use DCT-LMS algorithm with recursions on C to estimate optimum weight vectors for linear estimation.