A Study On Real-Time PIV Image Processing Teachnology

Hypersonic aircraft will play an important role in the field of military,political and economic.It has become a weapon for countries to take advantage of space strategy.As an ideal platform for studying on the measurement of hypersonic flow field,hypersonic wind tunnels are highly valued by various countries.Whether measuring techniques can accurately measure hypersonic flow fields has become the key of the development of hypersonic technology.Due to the unsteady nature of the hypersonic flow field,a great demands are placed on the measuring technology.Particle Image Velocimetry(PIV)has been widely used in high-speed flow field measurement.PIV involves PIV image processing algorithms and PIV image processing systems,and has both theoretical and practical significance.This article studies the PIV image real-time processing technology.The specific research work is as follows:(1)In view of the lack of real-time performance caused by the current PIV measurement technology using PC as a processing platform,this article studied the real-time processing technology of PIV images and developed a PIV image real-time processing system.The system can process flow field images in real time,the speed of image processing could achieve 100 fps,and the bandwidth can reach 10.24 Gbps.(2)For the real-time processing requirements of PIV images,a PIV image processing hardware platform based on FPGA+2DSP architecture has been designed.The platform uses the FPGA as the control core,the DSP as the processing core,and the display unit as the human-computer interaction core.Compared with traditional platforms,the transmission rate and processing rate are greatly improved..(3)According to the characteristics of PIV image processing hardware platform,the PIV image processing platform firmware system is designed.The system implements these parts: the reception of CameraLink data,the image data cache of DDR3,the configuration of the Switch chip,the transmission of SRIO high-speed data,and the transmission of Ethernet communications.Finally,the experimental verification of each firmware sub-module is carried out.The experimental results show that each sub-module works normally and the accuracy of data transmission is as high as 99%.(4)For the real-time processing requirements of PIV image processing system,pre-processing algorithm firmware system and flow field estimation algorithm firmware system are designed.The former uses Gaussian filtering algorithm to eliminate image noise,and the flow field estimation algorithm realizes real-time processing of cross-correlation algorithm and optical flow algorithm.Experiments are designed for the pre-processing algorithm firmware system and the flow field algorithm firmware system.The results show that the pre-processing firmware system can effectively remove the image noise.The flow field algorithm firmware system can calculate the velocity vector and amplitude vector of the flow field in real time.The processing speed is up to 100 fps.