Jul 08, 2015
Catchoom Team
Image Recognition, Image Recognition software
For the release of our Image Recognition Engine (codename CRE) version 3.3, we focused on improving the technology’s performance on recognizing objects in the presence of complex backgrounds.
One of the strengths of Catchoom products is that our team does its research and testing in scenarios where it’s most difficult to achieve successful results. This trial-by-fire methodology makes our products sturdier and more robust in situations where many image recognition technologies fall short.
We want Catchoom to become more and more easy to use. This means that we keep making improvements so that users can exert less and less effort, while experiencing even better results.
With this release of the CRE, recognition rates improve even in the presence of complex backgrounds. So if users are capturing a bottle label, for example, only 20% of the information in the image needs to be recognizable in order to achieve successful results.
The remaining 80% of the image can be fairly complex without affecting results. And you can bet that we will keep on improving this until Catchoom can handle any kind of background complexity!
Image Recognition Engine 3.3 also improves recognition of images in catalogs, newspapers, magazines or other printed media that are surrounded by dense text. In the past, text complexity may have impeded recognition results, but we’ve trained the engine to not be thrown off by dense text backgrounds.
Catchoom’s Image Recognition Solution already performs amazingly well under poor lighting conditions or with long distances to the object. These improvements to recognition in scenarios with complex backgrounds mean that you can deliver end-users even better results and greater ease of use.
Get our SDKs to enjoy your new power:
If you’re a CraftAR user, you’re already getting the benefit of this new release, since it’s at the core of the service. And we’ve already got our eyes set on the next tough scenario to tackle. Stay tuned!
