问题
The end goal would be to see if
contains
.
the compare needs to support minor distortion, scaling, color differences, rotation, and brightness differences.
it can be in any language really. i will be running this algorithm as a webservice so its no problem if i have to write this portion in c, c++, python, etc.
回答1:
You should probably take a look at OpenCV and VLfeat.
回答2:
Object detection can be performed for example using
- Rapidminer IMMI (image mining extension for one of the leading open-source data-mining platform)
- BoofCV (using SURF feature detection)
回答3:
How about ImageMagick? Its not a library per se however if you can provide shell access to your environment its pretty easy to use.
You would most probably be interested in the compare command.
EDIT: ImageMagick does contain tools for sub-image search like subimage-search.
回答4:
Look at this - http://gallery.azureml.net/MachineLearningAPI/02ce55bbc0ab4fea9422fe019995c02f - it supports OCR. This also supports multiple languages and distortion - http://www.projectoxford.ai/doc/vision/OCR
来源：https://stackoverflow.com/questions/6999920/looking-for-an-image-comparison-pattern-recognition-library