开始使用Open CV图像识别 [英] Get Started with Open CV image recognition
问题描述
我正在尝试使用Open CV创建一个用于图像识别的应用程序,我想实现类似)
如果形状发生变化,你应该看看更强大的方法,如SIFT或SURF 。两者都已在OpenCV中实现(来自aishack的链接是重新实现SIFT的教程,您可以在同一网站上找到使用OpenCV方法的教程)。
I am trying to make an app for image recognition with Open CV, i want to implement something like this but i don't know how should i do it can any one give me any help where should i begin from i have downloaded Opencv for iOS from here,
I have a hardcopy of image as an example which i want to scan through the camera and the images(markers) i have imported in project now when i scan the image through camera then it should overlay the markers on the image and when i tap/select the marker it should show the info of that marker. Here is my image :
It's just an example i have taken (Square,Circle and Triangle as Markers)
So now when the image is scanned then the markers will come up as an overlay and on clicking the markers i should get the names (If the Overlay image over the Circle Named "Air" is tapped it should show me "Air" on an alert or if Square Named "Tiger" is tapped it should say "Tiger") My problem is that the images are kind of same pattern but the result is different on every part so i don't know how should i approach in this ..
Please can any one help me out by suggesting any idea or if any one has done thing like this please tell me how should i implement it. I have to start from scratch any help please .
Can this be achieved using Open CV or i have to use any other SDK such as vuforia or layar.
Maybe you should search a little bit before asking help...
Anyway, the shapes you want to find do not seems to change (scale, rotation) so, you can look at the template matching methods implemented in OpenCV (see Tutorial OpenCV) If the shapes are changing, you should look at more powerful methods such as SIFT or SURF. Both are already implemented in OpenCV (the link from aishack is a tutorial to re-implement SIFT, you can find in the same website a tutorial to use the OpenCV method).
这篇关于开始使用Open CV图像识别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!