Data Mining Hardware Acceleration for Object Detection

Abstract

Over the past decade, object detection has been an active research area in the field of computer vision due to its many applications such as intelligent robots, smart homes, monitoring and surveillance. The need for real-time detection in most of these applications and the frequent failure of software-based implementations in achieving real-time capability, has motivated researchers to utilize hardware acceleration. Due to the potential for parallelism, low power and reconfigurability, Field Programmable Gate Array (FPGA) devices are well-suited for object detection system implementation. In this survey, we investigate FPGA-based implementation of object detection systems using data mining algorithms.