Identification of Air Pollution Levels Based on Sky Image Using Color Histogram Method

Air pollution is currently a widespread issue affecting various regions around the world, including Indonesia. This research aims to assess the extent of air pollution based on sky imagery. The study’s objective is to determine air pollution levels within specific areas using sky images and to develop software capable of implementing the research findings. The research comprises three main stages. The initial stage involves image input or extraction using the cropping method. The subsequent stage encompasses preprocessing, which includes three steps: converting RGB sky images to grayscale, enhancing image quality through contrast stretching, and reducing image noise using the median filter technique. The final stage entails image processing using the color histogram method, with the air pollution levels presented in the results section. The study employs a dataset of 20 sky images, comprising 10 images depicting good air conditions and 10 images representing moderate air conditions. These images are in.jpg format, with a pixel size of 1080 x 1350. The study’s results indicate that out of the 20 input images, 18 accurately reflect the air pollution level, achieving a success rate of 90%.