UncannyCV Android SDK

An optimized Computer vision Software Development Kit (SDK) tailored for the needs of an application developer.

An application developer targeting camera applications, often face performance bottlenecks whenever he tries to do more than just video recording. Most of the image processing operations are performance intensive. Most real time computer vision operations are not even feasible to do at good quality and high frame rates.

Here comes UnannyCV. We are built for ARM, the most ubiquitous computing platform of the embedded and smartphone/tablet world. Our rich set of algorithms help you overcome the image processing and computer vision performance bottlenecks.

SDK Contents:  
  1. UncannyCV library,pre-built for Android. This will work on all android devices that have an Cortex A series ARM processor. Your C code (using Android NDK) just needs to link to this library for calling the APIs.
  2. Test bench, contains the C test code invoking each of the above APIs. You can easily understand the usage of all of our APIs from this.
  3. Sample Android Camera applicationa full UI application utilising our algorithms. Source code for the sample is also included. You can just import this into Eclipse IDE and quickly get started.
  4. JNI glue, facilitates calling our APIs (which are built in Android NDK) from Java. This is a collection of Java files with specific package structure, which we use in our above Android application to call our C APIs. You can copy these files to your own Android application and invoke our APIs.
 
  1. You can see the detailed algorithm listing at Optimised Computer Vision Algorithms