A method for correction of rural multispectral aerial image mosaics

Aerial remote sensing of rural areas is increasingly used in forestry, precision agriculture, environment studies etc. The paper addresses a problem of creating visually acceptable multispectral mosaic images combined from data acquired in several flight lines. A method is proposed for the correction of brightness changes due to varying atmospheric conditions and influence of other factors. It is based on separation of image pixels related to vegetation and normalizing brightness using maximum values observed in green spectral channel of these pixels. It is shown that the visual appearance can be significantly enhanced in the proposed way and achieved quality of the image correction is similar in all parts of the mosaicked image. Correction of image pixels is performed in a way preserving ratios of intensity values in spectral bands.