ENHANCED AIRCRAFT DETECTION IN REMOTE SENSING IMAGERY THROUGH DEEP LEARNING ALGORITHM

Aircraft recognition in computer vision technology is essential, involving extracting the plane’s shape via a recognition processor. Image recognition, identifying objects in digital media, is pivotal in various applications, like automating industrial lines and security surveillance. Aircraft, with their diverse characteristics, pose challenges like varying size, texture, and coloration, compounded by disruptions like clutter and contrast differences. Neural networks, particularly convolutional neural networks, are utilized for recognition. The process involves median filtering for image processing, followed by feature extraction, utilizing shape, size, and texture. Superpixel segmentation reduces dimensionality, and convolutional neural networks classify using templates. An alert system notifies administrators upon aircraft detection, aiming for higher accuracy than current algorithms.