Abstract

A preprocessing methodology for the acquisition of quantitative relative reflectance data from multispectral multitemporal photography is proposed. The technique serves to suppress frame-to-frame and date-to-date variability, which influence the relationship between reflectance and photographic image tone, by the steps of gamma standardization, scene-to-scene ratioing and band-to-band ratioing. An example of the preprocessing of ground-level and aerial photography of four types of vegetation cover taken at regular intervals demonstrates the close covariance of the data sets with respect to each other and to seasonal variations in biomass and ground cover. The methodology has applications in all situations requiring quantitative relative reflectance data to be obtained economically from photographic data.