Criterion to Evaluate the Quality of Infrared Target Images Based on Scene Features
Keywords: Image analysis, image quality, object detection, object recognition
AbstractIn the field of infrared image signal processing, a novel criterion of evaluating infrared target image quality has been proposed in this paper based on analysis of the disturbance during target detection. To quantitatively characterize this criterion, two infrared target image quality metrics, namely IDGB (interference degree of global background) and SDLB (similarity degree of local background) are developed in this paper. Here, IDGB reflects the ability of the global background to disturb target search detection, and SDLB represents the degree of the local background to disturb target static detection. Experimental results show that the proposed metrics are more convictive than traditional metrics, such as SV, TSSIM, TTSIM.
Copyright terms are indicated in the Republic of Lithuania Law on Copyright and Related Rights, Articles 4-37.