31767687Kalman111

汽车牌照识别中车牌定位问题。车牌定位分为两步进行，首先对车牌进行初定位，对初定位后的车牌图像进行二值化、几何矫正，然后再精确定位车牌区域。同时，在matlab环境下仿真了汽车牌照识别，仿真结果表明，本文算法能很好的识别汽车牌照。(The problem of car license plate location. License plate location is divided into two steps, first at the beginning of vehicle license plate location, license plate on the positioning of the image at the beginning of two, the value of geometric correction, and then the accurate location of the license plate area. At the same time, in the MATLAB simulation environment of the vehicle license plate recognition, the simulation results show that, this algorithm can be a very good car license plate recognition.)

字符识别，MATLAB程序的。可以运行。基于神经网络(Character recognition)

Based on the rough set theory and its reduction methods, a MATLAB interface is written and included in the RSDA Toolbox

东北大学研究生上课时数字图像处理课的课件(Northeastern University graduate class of digital image processing course courseware)

主要用来计算给定观测序列的概率（求解HMM中的第一个问题）(Mainly used to calculate the probability of a given observation sequence (HMM solving the first problem))

用matlab中的GUI界面设计了一个可以实现混响的界面(Designed a reverberation of the interface can be achieved using the GUI interface in matlab)

几个非常实用的神经网络实例，是用matlab语言编写(Several very practical examples of the neural network, using matlab language worth learning)

说明：  均匀线阵上实现DOA估计，ESPRIT算法，16阵元（可调），半波长阵列流形。(DOA estimation on a uniform linear array, ESPRIT algorithm, 16 array elements (adjustable), half-wavelength array manifold.)

给出了一个简单的方法，使两个单元之间的联系的(gives a easy method of making a link between two cells)

用MATLAB编写的非近轴近似的高斯光束，可以更精确地描述高斯光束。(MATLAB prepared with a non-paraxial approximation of the Gaussian beam can be more accurately describe the Gaussian beam.)