Research on Image Processing Algorithm in High Resolution Infrared Imaging Based on Principal Component Analysis
Download as PDF
DOI: 10.25236/ictmic.2020.038
Corresponding Author
Fujun Wang
Abstract
High resolution imaging has always been the focus of research in the field of infrared imaging. It is not only of great significance in theory, but also has urgent needs in practice. The nonuniformity correction methods commonly used in image processing are analyzed and studied. Through analyzing the advantages and disadvantages of calibration-based and scene-based correction methods, a nonuniformity correction method of two points+high pass in time domain is proposed. The mechanism of non-uniformity is discussed from three aspects of infrared focal plane array, readout circuit and ambient temperature, and the exponential relationship model between detector response and ambient temperature is established. Aiming at the problem of detector nonuniformity drifting with time, three improved scene nonuniformity correction algorithms are proposed. Aiming at the contradiction between large dynamic range of signal source and small dynamic range of display output in infrared focal plane array signal processing, an adaptive histogram equalization technology for infrared images is proposed. The limitation of the dynamic range on the performance of the infrared focal plane array system is solved, and the image quality is improved.
Keywords
Principal component analysis, Infrared imaging, Image processing, Image interpolation