Research On Narrow Bandpass Filter Velocimetry Based On Area CMOS Image Sensor

The narrow bandpass filtering velocimetry system uses narrow bandpass filter with dark and white structure and outputs periodic signal whose frequency is proportional to the measured speed of the target. With combination of the spatial periodic signal and the time domain signal, measurement of displacement and velocity can be done. Such kind of system has good stability and simple structure without the use of coherent light source. Therefore it has shown its great potential and value in the field of velocity measurement of large size solid object in large space and long distance. As the optical spacing of electrophotonic detector decreases to the micron level, the velocity measurement method based on narrow bandpass filter is expected to reach higher measuring accuracy and gradually draws more attention in recent years and has futher development.Researches of this paper aim to independent displacement, velocity and related derivative geometric measurement with the random pattern distribution objects. Area CMOS image sensor is put forward as a narrow bandpass filtering component. New type of narrow bandpass filter is constructed with free pixles combination and weighed gray scale method. Thorough theory, simulation and experimental studies have been carried out. The main work contents and achievements are as follows:1. Methods of displacements and velocity measurement with high accuracy, as well as relative products, which are in populare use, have been studied. The type, characstics and developing trend of narrow bandpass filter adopted in narrow bandpass filtering method have been mainly analyzed. This thesis establishes the research on narrow bandpass filter velocimetry based on area CMOS image sensor for its several useful features such as high sensitivity, resolution, simple manufacturing process and simplicity, high collecting and reading speed and low power consumption.2. The mathematical model of the system has been established. Using time domain convolution and frequency domain spectrum operation, the transfer function of the system is derived based on transmission fuction of rectangle shape. Periodic structure with dark and white distribution can be treated as a narrow bandpass filter and can output periodic signal with spatial frequency of 1/. With the Combination with the spatial periodic signal and the time-domain signal, high-accurrace measurement of the speed of the object can be done. The two dimensional differential and direction identifiability CMOS narrow bandpass velocimetry structure diagram was given with the combination of differential demodulation technology, direction identification and two dimensional velocity measurement.3. The results of theoretical study and simulation computation of the six kind of narrow bandpass filters shows that:(1) All the six kind of narrow bandpass filters generate signal containing only the zero and 1/T frequencies;(2) The spectral width increases in the gaussion, rectangular and circle shape;(3) The spectrum based on sine function has no high frequencies, and smaller spectral width comparing with rectangular function;(4) If the number of cycles is larger than 10, the rectangular structure can meet the requirement, otherwise, the gaussion shape and sine function should be choose. Moreover, the optimal and failure models of the target are analyzed and factors which influenced the system performance are studied. The rules have been given about how to match the settings between the object’s measurability features, the optical magnification, and the signal acquisition speed.4. With the influence of the randomness of the objects, fluctuation of speed and light enviorenment, traditional counting methods based on spectrum scanner and frequency tracer are no more suitable for processing of the output signal whose characteristic in continuity, amplitude, and signal-to-noise ratio is random. Therefore, in this thesis, a method based on digital signal processing platform is proposed, adopted with related processing technology to suppress noise in a certain extent, Fourier spectrum transform and analysis, and linear spectrum correction technology. It is able to quickly get the signal frequency and do velocity measurement with high precision.5. Velocity measurement system with two dimensional differential direction identification narrow bandpass filtering technology based on CMOS detector is built for the targets consist of artificial random objects(mouse pad), decorative pattern of cement pavement and buildings(natural scenery). The results have shown that the displacement measurement accuracy can achieve 2% and 1.5% for targets with nonuniform contrast and continuous targets, respectively, with the proposed velocity measurement system. It even can be higher than 0.5% for anglar velocity measurement of targets with uniform contrast.The study of this paper shows that: The multiple narrow bandpass filters based on area CMOS image sensor whose pixels are freely combined can do high precision measurement on two dimensional displacement and velocity, without affection on other function of monitoring and measurement at the same time.