IMPLEMENTATION OF A REAL-TIME PROCESSING SYSTEM FOR OBJECT DETECTION IN DIGITAL VIDEOS USING FPGA
FPGA; Object Detection, Edge Detection; RTL Design; Functional Verification;
Digital video and image processing is an area that has many applications that need real-time performance, for example, Video Surveillance and Automotive Safety. A real-time image processing application aims to manipulate, analyze and process a large amount of data related to the pixels of an image in a stipulated time. These applications include image processing techniques like image analysis, image enhancement, image segmentation, object detection and others. Extracting the resources of an image is an intensive task with high computational cost and through the use of an FPGA device it is possible to achieve a more suitable implementation.
This work discusses the implementation of a Real-Time Processing System for Object Detection in Digital Videos using FPGA. All the steps of the FPGA design flow were developed. In this approach the steps of project specification, system modeling, microarchitecture, RTL coding, functional verification, logic synthesis, gate level simulation (logic post-synthesis), time analysis and FPGA device programming were included.
The proposed hardware architecture was designed and synthesized for Altera DE2-115 FPGA kit. The results are verified in real time with an input video from TRDB-D5M Terasic CMOS sensor, DE2-115 FPGA kit and VGA monitor for displaying images.