TY - JOUR AU - T. N, Padmini AU - T, Shankar PY - 2022/10/01 Y2 - 2024/03/29 TI - Static Weather Image Classification Based on Fog Aware Statistical Features Using XGBoost Classifier JF - International Journal of Intelligent Systems and Applications in Engineering JA - Int J Intell Syst Appl Eng VL - 10 IS - 3 SE - Research Article DO - UR - https://ijisae.org/index.php/IJISAE/article/view/2140 SP - 64-74 AB - <p>The outdoor functions carried out by the autonomous navigation systems fail in extreme weather conditions. To tide over such issues, many researchers have implemented an algorithm to get rid of the fog, rain and snow from images. Most of the dehazing algorithms are implemented by researchers considering the input image as a hazy image. But in the real-time scenario, the image captured by the camera can be any image with or without degradation due to the influence of the weather. This research is a proposal to classify static weather images like haze and fog along with sunny images using a supervised classifier. It can be stated that this is a pioneering opportunity to analyze fog and haze as two separate classes. Other researchers have hitherto treated them as just one class. The proposed method was implemented by collecting images from existing databases and forming a new database by relabeling the images as haze and fog based on psycho-visual analysis. The classification model was trained and tested on static weather images using a supervised classifier. It was inferred that the XGBoost classifier has a definite edge over such other classifiers in existence.</p> ER -