Analysis of small target detection algorithm based on image gray entropy

The local entropy algorithms (LEA) for aerial target have been introduced in the past, which were focused on its applications and improvements. In this paper, by the research and analysis about the features of the local entropy in detail and deeply, the relationships are studied quantitatively for the variations of the local entropy values for the single pixel target with the different contrast and the effects to this target from the different sizes of the neighbors. Also, the relationships between contrast, k, target size, m, and neighbor size, M, in local entropy, H(u,v), are analyzed. These analyses provide the quantitative analysis base for the aerial small target, such as infrared targets.