Analysis of video data from an unmanned aerial vehicle based on a structural similarity index

Abstract

The article proposes a metric for the analysis of video data recorded by an unmanned aerial vehicle that uses a structural similarity index for evaluation. The metric consists in comparing frames in terms of brightness, contrast and pixel structure and a subsequent assessment of the video frame state. A comparative analysis of the proposed and currently employed metrics was carried out. The research included simulations on analog and digital video data at different frame rates. The results showed that the developed metric successfully detects delays, frame distortions and dynamic changes in a video scene. The proposed metric can find a wide range of applications of unmanned aerial vehicles in applied areas: construction, agriculture, geology and cartography.