Imagery recorded using satellite sensors operating at visible wavelengths can be contaminated by atmospheric haze that originates from large scale biomass burning. Such issue can reduce the reliability of the imagery and therefore having an effective method for removing such contamination is crucial. The principal aim of this study is to investigate the effects of haze on remote sensing imagery and develop a method for removing them. In order to get a better understanding on the behaviour of haze, the effects of haze on satellite imagery were initially studied. A methodology of removing haze based on haze subtraction and filtering was then developed. The developed haze removal method was then evaluated by means of signal-to-noise ratio (SNR) and classification accuracy. The results show that the haze removal method is able to improve the haze-affected imagery qualitatively and quantitatively. © 2019 International Journal of Advanced Computer Science and Applications.