基于matlab的指纹识别系统，采用基于细节点对比，输入图像必须是256×256，8－bit灰度值，500dip的指纹。...

利用PSO优化SVM，利用分组式训练方法提高算法速度-PSO to optimize the use of SVM, the use of packet-style training methods improve algorithm speed

基于EKF的机器人SLAM演示程序, 对于搞机器人定位及地图构建的朋友会有一定帮助, 这个版本用的人比较多-EKF-based SLAM demo of the robot, the robot carry out positioning and mapping will have a certain friend to help, this version used by many people

two Matlab functions for initializing and training a recurrent neural network -two Matlab functions for initializing and training a recurrent neural network

遗传算法在工程优化计算上的应用的MATLAB程序。-genetic algorithm optimization calculation of the MATLAB program.

模拟退火算法求解经典图论中的图着色问题的源程序-simulated annealing algorithm classic graph coloring map of the source

模糊数学与人工智能技术 -fuzzy mathematics and artificial intelligence technology

利用Fisher分类器对数据进行分类，并画出决策面（含测试数据）-Fisher classifier using the data classification, and face painted in decision-making (including test data)

c语言实现人工智能神经元网络boltzman算法源码-c language artificial intelligence neural network algorithm source Boltzman

停车诱导系统中车位预测模型的研究 摘 要 研究城市停车诱导系统的停车车位占有率预测问题。首先提出墓于B P神经网络的车位占有预测模型， 同时将自适应 学习速率调整法和加入动量项方法用于改善基本B P神经网络， 优化了学习速率， 减少了训练过程的震荡趋势， 改善了网络的收效 隆。以此为基础实现了停车位的智能预测 0最后， 进行了多种方法比对实验-err

基于canny的灰色预测和模式识别,标定的一种做法是：在一块平板上画若干个点， 同时用这两部相机照相，分别得到这些点在它们像平面上的像点，利用这两组像点的几何关系就可以得到这两部相机的相对位置。然而，无论在物平面或像平面上我们都无法直接得到没有几何尺寸的“点”。实际的做法是在物平面上画若干个圆（称为靶标），它们的圆心就是几何的点了。而它们的像一般会变形，如图1所示，所以必须从靶标上的这些圆的像中把圆心的像精确地找到，标定就可实现。-Based on the gray canny prediction and pattern recognition, calibration approach is to: a plate painted in a number of points, at the same time with these two cameras camera, respectively, as these points in their plane as points, the use of these two groups geometric relationship between the image point can be that the relative positions of two cameras. However, in terms of objects such as plane or plane, we can not directly be no geometry point. Actual practice is complex plane painted in a number of round (known as targets), their geometric center of a circle is the point. And they generally like deformation, as shown in Figure 1, it must target these round, like in the center of a circle of like to find precise, calibration can achieve.