2022 № 2 Сучасні інформаційні системи
Постійне посилання колекціїhttps://repository.kpi.kharkov.ua/handle/KhPI-Press/65928
Переглянути
Документ Methods of compensation of microbolometer matriсes self-heating in the processing of thermal images(Національний технічний університет "Харківський політехнічний інститут", 2022) Zuev, Andrey; Ivashko, Andrey; Lunin, DenisThe sources of noise and artifacts arising during thermal imaging and the methods for thermal images filtering, including methods specific for processing of images generated by infrared sensors, are considered. In particular, distortions caused by the process of microbolometrer matrices self-heating due to internal and external heating sources and the methods for compensating such distortions are studied. The purpose of the study is to create a mathematical model of a bolometric matrix self-heating based on heat transfer equations and to develop an algorithm for suppressing of distortions introduced into thermal images by self-heating. The exponential models describing the propagation of heat in the microbolometer matrix are proposed and it is shown that the coefficients of the models after logarithming can be determined by the least squares method. For real thermal images, the coefficients of the model are determined, and situations are considered when the base temperature of the object is known and when it is necessary to restore it, and modifications of the exponential model in the form of an exponent from a complete and incomplete square are proposed. Computer simulation of the proposed distortion compensation algorithm has been carried out, a set of thermal images before and after processing has been presented, and a quantitative estimation of the degree of noise suppression caused by heating of bolometric arrays has been obtained. Based on the results of the work, it was determined that the exponential model provides a sufficient degree of closeness of the experimental and theoretically predicted temperature data, and the degree of difference between the data and the model was estimated. Recommendations are developed for the application of the proposed methods at known and unknown base temperature of the matrix. Proposals have been developed for further improving the mathematical model, including the situation of temperature changes over time, and for improving the efficiency of self-heating noise suppression algorithms.