Research On The Key Technologies Of Two-color Infrared Imaging System

Temperature is among the most crucial parameters for characterizing the material status. Measuring and controlling temperature is closely related to the safety, efficiency and quality of production, as well as energy conservation. Infrared imaging temperature measuring technology has been rapid developing and widely used in industrial production due to the advantages of non-contact detection and high precision. Because of the material, technology and target emissivity, the issues of non-uniformity and measurement precision need to be solved for infrared thermal images. A new dual-band emissivity compensation algorithm for target is developed. And dual-color infrared imaging temperature measurement equation is given.Aiming at the issues of non-uniformity of focal plane array(FPA), the causes were analyzed. And the linear and nonlinear responding models of the FPA were introduced. Furthermore, several FPA non-uniformity correction algorithms(e.g. two-point algorithm, multipoint algorithm and the polynomial algorithm) were introduced in detail. In this study, we developed a correction algorithm on the basis of nonlinear responding model. Computing the output deviation between the actual output pixel value and the expected value, a standard least square curve fitting was performed with a polynomial to build the nonlinear model between the deviation and the output value. The results indicated that the non-uniformity issues were well addressed in this study.Due to various causes, the FPA has a lot of blind-pixels. The potential causes of blind-pixels were discussed and several detection methods including those based on definition, background and blackbody, as well as 3? and window were introduced. This paper also introduced a variety of blind-pixels compensation algorithms including linear interpolation method. Finally the results of blind-pixels detection and compensation based on 3? and window detection algorithm and neighborhood compensation algorithm were acquired. The results showed that the proposed blind-pixels problem has been solved in good quality.This paper also carried out an accurate temperature measurement research using infrared thermal images. The basic equations of infrared temperature measurement were given, after which the basic theory of infrared radiation was introduced. According to the colorimetric temperature measurement method, the equation of the dual-band temperature measurement equation was derived. In consideration that the emissivity which can affect temperature measurement accuracy is difficult to obtain, temperature ratio approximation to target emissivity ratio in dual band was proposed to compensation target emissivity. Finally, the laboratory experiment of temperature calibration was conducted, which establishing the relationship between calibration source infrared imager dual-band output digital values and corresponding temperature.