Object Detection using a Novel YIQ Model based Image Fusion for UAV Aerial Surveillance

摘要

Multi-sensor image fusion is an important precondition of realizing target perception for unmanned aerial vehicles (UAVs), and then UAV can execute various given missions. Imaging devices on UAV are used to capture visual scenes of the ground and they are fused to extract the information. Both the imaging devices setup in the UAV and the target in the ground are in dynamic environment. Due to the imperfection of imaging device and instability of the observed scene, captured images often blurred and exhibit unsatisfactory spatial resolution. Hence single sensor does not fulfil the visual surveillance need in daytime. This paper propose the concept of fusing multi sensor aerial images captured by UAV using YIQ colour model based on Laplacian pyramid transform with Principal component analysis technique. The performance of the proposed image fusion scheme is analyzed using the evaluation metrics of Entropy, Mean Square Error (MSE) and Peak Signal to Noise Ratio (PSNR) with the existing techniques.