Future of the Augmented Reality

Augmented reality technology grows rapidly for last two years. The huge potential this technology has not fully revealed. In the near future we expect appearance of large number of companies seeking to take a a new area in the market. More and more quality and exciting applications of augmented reality will appear. Want to know why?

A very fast BGRA to Grayscale conversion on Iphone


Almost all image processing algorithms uses gray scale images as input source. But almost all hardware video sources provide frames in RGB/BGR(A) formats. So gray scale conversion is very popular operation. Although it’s expensive enough to cause CPU-bound bottlenecks while running on mobile processors. In this post i will show you how to use ARM NEON intrinsic to get significant performance boost of BGRA to GRAY conversion.

Markerless Augmented Reality on iPhone

Hello everyone! Today i want to share my results in research of markerless augmented reality. The  main idea - do fast and quality AR without those damn markers and give the ability to use real object as a target. Markerless augmented reality is very similar to marker-based systems like ARToolkit with one major difference - such technology use real object as a target for augmentation. It can be almost any kind of objects - photos, logos, beer bottle or Cola can.

Comparison of feature descriptors

Image for attraction of your attention Hello everyone! Today, we have very interesting topic! We will inspect different feature descriptor extractors. From this post you will know how robust is SURF, which disadvantages has BRIEF descriptor and how many times LAZY descriptor is faster than SURF. PS: I will be really appreciate if you point me to good implementations (C/C++) of RIFF, PCA SIFT, GLOH, LESH descriptors. I will include them in test suite.

Feature descriptors: A new approach

Last year I was tightly connected with image processing and feature tracking/matching. For my needs I’ve used SURF and later RIFF descriptors. Both of them have strong advantages and but… SURF descriptor robustness are compensated by it’s computational cost. RIFF descriptor extracts much faster but not robust enough for my needs. My needs are very simple – doing markerless AR on mobile phone.