BP网络进行0_9十个数字图像的识别
·BP网络进行0_9十个数字图像的识别...
·BP网络进行0_9十个数字图像的识别...
首先对图像进行灰度化、二值化、平滑去噪、归一化、细化的预处理,对设计好的神经网络进行训练,对比训练的结果与期望的结构,并根据对比的结果对神经网络的一些权值进行修改,最终得到训练好的神经网络。选择测试样...
目前,在字符识别研究中使用最广泛的是光学字符识别,即OCR 方法。其统计模式识别方法注重数量特征,便于特征提取、分析和计算。但是,它将字符看成一种随机的二维点阵,没有考虑字符的结构特征和结构信息。因此...
this tar includes my code which employ the Lin-Kernighan algorithm to address the tsp problem. this ...
aPLib is a 32-bit compression library based on the compression algorithm used in aPACK (my executa...
AC Induction Motor Control Using the Constant V/f Principle and a Space-vector PWM Algorithm...
·期刊论文:Image Segmentation Using an Improved Watershed Algorithm...
·国外论文:Image Segmentation Using an Improved Watershed Algorithm...
资料->【E】光盘论文->【E1】斯坦福博士论文->00 Texas-Austin PhD New algorithm for on-the-fly ambiguity resolution of r...
ucos From a small number of the concept of real-time kernel, developed a simple algorithm available,...