Use of local fuzzy variance to extract the scattered regions of spatial stray light influence in hyperspectral images

Spatial stray light is a phenomenon caused by scattering of photons from target objects, and it is generated within and among adjacent objects. The intensity and radiometric composition of spatial stray light depends on the light source properties and chemical and physical properties of the objects being imaged. A special kind of region exists in hyperspectral images, referred to as the “scattered region” of spatial stray light, which can be regarded as a boundary between different objects or between an object and background. In this study, local fuzzy variance was used to extract the regions, in which spatial stray light may have significant influence. The proposed method was applied to three data sets: synthetic array data and two hyperspectral imaging datasets (1: plastic card and 2: maize kernels from three inbred lines). Using local fuzzy variance method, scattered regions of spatial stray light could be extracted and visual inspection of data sets demonstrated the effectiveness and sensitivity of the proposed method.