Aircraft detection using single shot scale-invariant face detector with viridis saliency map from remote sensing image

Abstract

Aircraft detection from remote sensing images is a useful task and essential in many fields; for example, in the military, aircraft detection systems are used to search and inspect an enemy aircraft from remote sensing images. Nowadays, aircraft detection from remote sensing images has gained increasing research attention because this task is very challenging. This task is difficult due to many occurred problems in remote sensing images, such as a shadow, a complicated background, and various aircraft scales. In this research, we create a Viridis saliency map using many techniques from image processing such as color mapping, morphological image processing, and image thresholding. The Viridis saliency map can help reduce the noise from the background of remote sensing images well. Moreover, we combine our Viridis saliency map with a famous detector called Single Shot Scale-invariant Face Detector, which is used to detect a human face in the image. From the results, this method can handle the scale problem efficiently.