PIXEL QUANTIZATION INFLUENCE ON REAL-TIME CROSS-CORRELATION COMPUTATION FOR AN EMBEDDED PIV CONTROL SYSTEM

摘要

The main purpose of this paper is to evaluate the measurement efficiency of the PIV (Particle Image Velocimetry) technique in real time. Thus, we propose a new approach using pixel quantization format reduction to process image cross-correlation as well as flow displacement vector validation in real time. To achieve this, two delayed images are acquired and parallely transferred to a hardware system based on FPGA and to a computer. A cross-correlation processing algorithm between the images leads to computing of the displacement vector field. The PIV data validation criteria will be detailed and some of them are selected for implementation. With this contribution hardware needs are slightly dwindled by maintaining the same performance.