|
Infrared object detection is an important technique of digital image processing. It is widely used in automatic navigation,
intelligent video surveillance systems, traffic detection, medical image processing etc. Infrared object detection system
requires large storage and high speed processing technology. The current development trend is the system which can be
achieved by hardware in real-time with fewer operations and higher performance. As a main large-scale programmable
specific integrated circuit, field programmable gate array (FPGA) can meet all the requirements of high speed image
processing, with the characteristics of simple algorithm realization, easy programming, good portability and inheritability.
So it could get better result by using FPGA to infrared object detection system.
According to the requirements, the infrared object detection system is designed on FPGA. By analyzing some of the
main algorithms of object detection, two new object detection algorithms called integral compare algorithm (ICA) and
gradual approach centroid algorithm (GACA) are presented. The system design applying FPGA in hardware can
implement high speed processing technology, which brings the advantage of both performance and flexibility. ICA is a
new type of denoising algorithm with advantage of lower computation complexity and less execution time. What is more
important is that this algorithm can be implemented in FPGA expediently. Base on image preprocessing of ICA, GACA
brings high positioning precision with advantage of insensitivity to the initial value and fewer times of convergence
iteration. The experiments indicate that the infrared object detection system can implement high speed infrared object
detecting in real-time, with high antijamming ability and high precision.
The progress of Verilog-HDL and its architecture are introduced in this paper. Considering the engineering application,
this paper gives the particular design idea and the flow of this method's realization in FPGA device. And we also discuss
the problems on how to describe the hardware system in Verilog-HDL. Based on the hardware architecture of infrared
object detection system, the component units of the system are discussed in detail, such as image data acquisition unit,
data pre-processing unit and logical control unit etc. The design of the FPGA function and its implementation are carried
on Verilog-HDL with TOP-DOWN method. The ending is the prospect of the project.
|
