Under the effect of solar variation, atmospheric attenuation and thermal radiation distribution, the grey value of interference source is close to or equal to the target grey value. With the distance between the imaging system and the target being very long, the target is dotted or mottled in the image. In addition, the target motion is usually unknown and abrupt. All these factors cause the negative effect of small target detection. In this paper, we present a motion estimation and spatial-temporal filter-based algorithm to detect infrared small targets under complex background. The algorithm is divided into three steps: 1 researching the image transformation technology from greyscale pattern to entropy pattern; 2 studying the anisotropic smoothing and local constraint criterion of entropy flow; and 3 introducing spatial-temporal filter for detecting small targets through energy accumulation and trial search of targets. Compared to other methods, experimental results indicate that the proposed method can robustly detect small targets under complicated background.
Read full abstract