Design Of Embedded Image Acquisition Device And Its Application In BOS Mobile Display

Background Oriented Schlieren(BOS)is a novel flow visualization technique,which developed on traditional Schlieren technology.It has been widely applied due to its advantages of simple device structure,wide field of view,high spatial and temporal resolution.This paper use colored-fringe as background image,mainly focuses on following aspects:1.This paper designed and implemented an embedded image capture device based on FPGA,which has potential to be applied to actual colored-fringe BOS system for flow visualization.Because the process of image capturing in traditional BOS system is based on PC,power consumption and volume are large,the cost is high and the customization is difficult.Traditional BOS system has low resolution and limited capture speed in flow visualization.It is usually unable to observe flow field in real time,which is not convenient for system construction and adjustment.While embedded system and FPGA have many advantages such as low cost,small power consumption,programmable,easy to customize.Altera Cyclone IV EP4CE6F17C8 N FPGA is used to control core system.OV5640 CMOS image sensor is used to capture high-definition images with high-speed,with 30 FPS,1024*720 resolution,and RGB565 format.The captured image data is stored in SDRAM and can be displayed in real-time through VGA interface.There are three display modes:color,gray,and edge detection.Aiming at embedded image capture device based on FPGA,firstly,design system hardware framework and complete device selection.Altium Designer16.1 was used to complete schematic design and system PCB design.Then,design logic of various modules by Quartus Prime 17.1 environment and Verilog HDL.Finally,complete circuit and system debugging in laboratory,given the final device result.2.This paper use colored-fringe pattern as background,analysis BOS background image processing method based on three-step phase-shifting algorithm.one of the crucial steps of BOS technique is selection of appropriate background image and method to extract image displacement vectors.At present,the algorithm of displacement extraction in BOS technique usually has cross-correlation algorithm,optical flow algorithm and dot matrix recognition algorithm.However,these algorithms have disadvantages such as difficulty in adaptive size of cross-correlation interrogating window,weak anti-noise ability and low spatial resolution.To solve this problem,this paper proposes a BOS image processing method for background image based on three-step phase-shifting algorithm,using colored-fringe pattern as background.The RGB coded carrier-fringe image in our method is composed of three phase-shifted images,replacing the printed or drawn speckle pattern in traditional BOS method.Displacement vectors are extracted by comparing different phases between the corresponding points in three separated images,which avoids the problem of selecting interrogating window.The flow visualization experiment is carried out by use of the proposed colored-fringe BOS method based on Three-Step Phase-Shifting,and the experimental results are analyzed to verify the feasibility of proposed method.