Research On Image Acquisition Of Fiber Optic Accelerometer Based On OV9121

With the rapid development of aviation, aerospace, construction, machinery and other areas, the vibration problem got more and more attention. Accordingly, higher requirement for the performance of the acceleration sensor has been made, and it is the current developing trend to get acceleration information quickly and at real time. This paper proposes image sampling method based on an integrated fiber-optic acceleration sensor. By using this method we can make the acceleration sensor system achieve the high-speed data acquisition of image, to realize real-time measurement target. It also has advantages of moderate light coherence demand, and little influenced by environmental temperature change etc.The acceleration sensor system given in this paper mainly consists of the following four parts as the sensing probe, the light source constant current drive, the CMOS image sensors and the image signal processing. As a sensing element, the twin-core fiber is bent by inertial force. The light constant current drive provides a stable optical power injection for the system so as to obtain high quality images. CMOS image sensor is the image receiving device. Image acquisition and processing adopt DSP + FPGA architecture to realize embedding and real-time.From the start of accelerometer, this system has implemented the design and the production of sensor probes first, and the length of the sensor head can be adjusted to accommodate changes in the measuring range; and then a light constant current drive circuit is completed, which can achieve the adjustable scope within range 0-50mA so as to satisfy to the experimental conditions and device requirements. Using OV9121, the CMOS image sensor integrated AD transform module inside the chip, and output 10bit grayscale data.FPGA chip completes the SCCB bus driver which realizes the changes of the video window and the frame frequency, meanwhile, through adopting median filtering algorithm, FPGA chip realizes the image collection and preprocessing, completes the system integration and real-time measurement, thus facilitates the obtainment of acceleration value by completion of DSP chips fringing shift amount calculation.Eventually, its software and hardware debugging is completed, verifying the feasibility of the system. Finally, the improvement plan has been put forward to the problem, which will show the direction for future project research.