使用opencv在iPhone中进行图像匹配/检测 [英] image matching/detection in iphone using opencv

查看:90
本文介绍了使用opencv在iPhone中进行图像匹配/检测的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!

问题描述

我有2张图片,例如bigImage和smallImage.我想检测smallImage是否在bigImage中的任何位置,而不论其方向或transforms(rotations).如果存在,则应返回true,否则返回false.我已经经历过openCV中的模板匹配(cvMatchTemplate)方法,但是由于对与错匹配的输出之间没有太大差异,因此没有到达任何地方.使用cvMatchTemplate是否仍然可以满足我的要求,或者在openCV中还有其他方法可以做到这一点.

I have 2 images, say bigImage and smallImage. I want to detect whether the smallImage is there anywhere in the bigImage, irrespective of its orientation or transforms(rotations). If its there, it should return true and otherwise return false. I have been going through the template matching (cvMatchTemplate) method in openCV, but haven't reach anywhere since there aren't much difference between the output for a true and a false match. Is my requirement still possible using cvMatchTemplate or are there any other methods in openCV for getting to this.

推荐答案

这是一个非常普遍且非常棘手的问题,有很多解决方案.它们都不是银弹.希望其中之一将为您解决特定问题.尝试看一下SIFT,颜色直方图匹配,SURF,HoG,Viola-Jones,DOT和ORB.其中一些具有OpenCV实现;其他人没有.您应该阅读可用的选项以及它们的优缺点,然后尝试一下.

This is a very general and very hard problem, with lots of solutions out there. None of them are silver bullets; hopefully, one of them will work for your particular problem. Try taking a look at SIFT, color histogram matching, SURF, HoG, Viola-Jones, DOT, ORB. Some of these have OpenCV implementations; others don't. You should read up on the available options and their strengths and weaknesses, and then try some out.

这篇关于使用opencv在iPhone中进行图像匹配/检测的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!

查看全文
登录 关闭
扫码关注1秒登录
发送“验证码”获取 | 15天全站免登陆