Abstract This paper describes and evaluates a number of techniques for reducing different types of
noises which associated with the thermal images. These techniques are based on optical image
filtering in both spatial domain and frequency domain. Filtering in both spatial domain and
frequency domain are applied on different thermal images associated with three standard
noises models encountered in most images as additive, multiplicative, and impulse noises
with different variance. Also, Non-uniformity correction techniques are applied on several
thermal images associated with Fixed Pattern Noise (FPN). The algorithms have been tested
by using several real image data from existing infrared imaging systems with good results.
Measuring criteria for performance evaluation of thermal images enhancement techniques as
Peak Signal-to-Noise Ratio (PSNR), Signal-to-Noise Ratio (SNR) and Root Mean Square
Error (RMSE) are used to ensure the vision observation of user to select the most suitable
technique with highly performance evaluation.