Inspection on Infrared-Based Image Processing

Abstract

Infrared radiations have a large wavelength of 0.7 μn to 1mm compared to the visible light wavelength. All objects above 5 degrees temperature emit IR radiations. It is not possible to see the IR wave directly from the human eye without any aided sensing device. Infrared images contain insufficient contract and having a high rate of background noise which makes it difficult to view the captured image clearly. There are several factors (e.g. background nature and clutter characteristics etc.) based on which the performance of the imaging system gets affected. The most important parameters for the quality of the infrared image are the radiation pattern and the temperature difference between background and object. In this paper, inspection has been performed on Infrared (IR) image processing. A detailed study can be seen on IR detectors, a difference in visible images and IR images and on methods for non-uniform correction in IR signal processing. Theirs exist a variety of algorithms for the infrared image processing, some of which that are discussed in this paper are wavelet transformation, median filter, frame averaging, and spatiotemporal filtering. There are many other IR image processing algorithms which are used according to the specific requirement, to enhance the efficiency of IR image processing system, there is an extreme requirement of adaptive algorithms so that an optimum result can be achieved for IR image processing. Similarly there exists multiple IR imaging systems which are still used in different scenarios of IR image processing.