weixin_37943144的博客

One-Class SVM（一分类支持向量机）,vs版本 如果出现全-1 0 1的情况，且秒出训练（没有出现省略号）的同学，应该是代码有误，可对比参考； 文中使用boost库拆分文字，无boost库的同学可以自己修改一下 If there is all return - 1，0，1 . and the second training (no ellipses) , should be the wrong code, can be compared to the reference; In this , use the boost library split text, no boost library can change it

c++ BMP格式图片.不用使用opencv等库,提取,转化为BGR. struct Piexl24 { unsigned char blue; unsigned char green; unsigned char red; };