Image Processing And Experimental Measurement Of A Doppler Global Velocimeter System

Doppler global velocimetry (DGV) is a recent developed techniquebased on molecular/atomic filter principle and laser Doppler shiftphenomenon. Smaller tracing particle can be used in DGV technique, so it issuitable for the cases of hypersonic flow and large-size wind tunnelmeasurements. Since data processing method is simple for DGV, real-timemeasurement can be achieved easily. The main purpose of this paper is tostudy the image processing method of a Doppler global velocimeter, and todevelop a computer program for DGV image processing. In the study, a DGVsystem will be developed, and measurements will be made on it. Theperformance of the DGV system, including hardware and software, and thedeveloped image processing program will be evaluated based on themeasurement results.In the thesis, the background of DGV technique and its developinghistory were introduced, and some important achievements in DGV researchworks were stated. For the purpose of better understanding to this technique,the working principle and typical system of DGV were elaborated.The most important work in this thesis is to study the DGV imageprocessing methods, including the development of relevant computerprograms. In the DGV Image processing, three manipulations must be made,including background removal, gray scale normalization and geometriccorrection. By background removal, scattering and reflection of theenvironment light can be filtered out, thus the interference coming from thesurrounding environment background can be effectively reduced or removedout. By gray scale normalization, the difference between the two cameras andthe optical paths will be eliminated. By image geometric correction, geometrydeformation of a flow field image, which is unavoidable when using a camerato catch a picture of an object, can be corrected. In the study, an operating software was developed for the DGV system,and the computer programs for image and data processing were developedalso. The test results show that the software and the programs worked well.Experimental measurements were carried out on the optical elementsperformance, calibration of the iodine filter, and the velocity distribution on arotating wheel. The effects of temperature on absorption curve of the iodinefilter have been examined.In the rotating wheel experiment, peripheral velocity varied from6m/s to30m/s. The results show that the maximum error is less than5m/s and RMSvalue of velocity bias is small than2m/s, the velocity distribution obtained byDGV is consistent with the actual value, which indicates that the developedDGV system is reliable and the image processing computer program is usable.Comparing the results while the images were saved using8-bit and12-bitdata formats, it’s found that the higher gray-resolution image is helpful toimprove the measurement precision. In the end of the thesis, uncertainty ofthe measurement using this DGV system was analyzed, and discussion hasbeen made in terms of several possible error sources. The measured resultsand the discussions are valuable for the DGV system improvements.