OpenCV library is widely used by computer vision engineers across the world. It contains almost all algorithms you may want for R&D or product development. It has production-ready build farm with tests and strong community that give nice feedbacks and discover errors. But nevertheless OpenCV has some strange issues and undocummented behaviour that can surprise you as minimum and crash your app as maximum.

Recently i came across the publications to a new features called KAZE (Japanesee work meaning “Wind”). They interested me, because KAZE authors provided very promising evalutaion results and i decided to evaluate them too using my OpenCV features comparison tool. Fortunately KAZE algorithm is based on OpenCV, so it was not too hard to wrap KAZE features implementatino to cv::Feature2D API.

With this post i start describing my experience with a “Image and video processing: From Mars to Hollywood with a stop at the hospital” class. I will write a series of 9 posts each week i complete and highlight interesting tasks and solution to optional programming quizzes in OpenCV.

This is a second part of “Image and video processing: From Mars to Hollywood with a stop at the hospital” class.

In optional taks for Week 2 we were asked to test various predictors for JPEG-LS compression algorithm. In this post i want to share my results and discuss how we can achieve better prediction rate.

During my research and development work i often have to display a lot of text infromation on top of the OpenCV images. You know what i mean. Suppose you’re writing video stabilization algorithm. On each frame you want to display number of features visible on current frame, number of features matched with previois frame, camera motion parameters, recent twitters, name of your pet, etc.. In the OpenCV you can use cv::putText function to print formatted std::string at the desired position on the image. It’s easy when you have a fixed string that won’t change. But what should you do if you need to dynamically construct this string depending on the processing results?

I feel excited writing this post. The “Mastering OpenCV with Practical Computer Vision Projects” books is done and published! This is my first experience as a book author and i hope you will like it. This book has been written by several authors, and covers many topics that has real appliance in computer vision area: augmented reality, face recognition and head pose estimation, structure from motion estimation and cartoon image processing. We as authors made our best to bring the most popular and demanded topics to you.