GPU Computing Gems Emerald Edition

Chapter 33. Haar Classifiers for Object Detection with CUDA

Anton Obukhov

This gem covers aspects of the approach to object detection proposed by Viola and Jones [1] . The algorithm is discussed along with its well-known implementation from the OpenCV library. We describe the creation process of the GPU-resident object detection pipeline step by step and provide analysis of intermediate results and the pseudo code. Finally, we evaluate the CUDA implementation and discuss its properties.

The gem is targeted at the users and programmers of computer vision algorithms. The implementation and analysis sections require basic knowledge of C/C++ programming languages and CUDA architecture.

33.1. Introduction

Object detection in still images and video ...