Implementation of fire image processing for land fire detection using color filtering method

Abstract

Peat swamp forest fires are the process of burning widely dispersed organic matter, which can be caused by burning waste, peat, grass, twigs, weeds, shrubs, leaves and fresh trees. Forest fire monitoring can be done using satellite patterns or unmanned aerial vehicles (UAVs) with pattern modeling applications. The image is interpreted by a computer so that it can be understood by humans. The purpose of this study is to implement the color filtering method by determining the value (HSV) and thresholding value of the image of the fire on peatland. the image will be classified to obtain environmental pattern recognition data. The results obtained are a value of fire images in HSV. HSV value for HEU is in the range 0-18, saturation value: 75-166 and value: 200-230. While the error test results for every 46 HSV pixel values are 4, so the percentage of errors is 8.89%.