Based On Ambient Temperature Thermal Imager Calibration And Hardware Implementation

The original image made by thermal imager always the problems of the non-uniformity associated with environmental temperature, blind pixels, and the horizontal or vertical stripes which affect the picture quality. For these problems, correction methods for each problem are proposed to achieve thermal integrated calibration based on the environmental temperature. In order to improve the calibration accuracy, the new non-uniformity calibration method based on "S" model is proposed after a detailed analysis of the causes of non-uniformity correction and many commonly calibration methods used before, besides, the influence of environmental temperature is considered based on the analysis of experimental data, the model of environmental temperature on the non-uniformity is established, according to which, the environmental temperature pre-compensation algorithm is described. In order to improve the pixel detection accuracy and make up the shortcomings of the fuzzy image detail by common blind pixel detect and compensation methods, after the analysis the cause of blind pixels and the common detect and compensate methods, the edge detection technology is brought in so that the new blind pixels detection and compensation algorithm based on keeping the details is proposed; According to the feature of horizontal and vertical stripes in the frequency domain, the horizontal or vertical correction by FFT is proposed,besides, the correction methods by wavelet transform is also proposed to lower detail damage of image by FFT. The DSP system based on thermal imager integrated correction algorithm is introduced, and the flow chart of algorithms mentioned above in the DSP system is described. It's proved by experiments results that the quality of the original image by the thermal imager is improved after the integrated correction based on environmental temperature.