The availability of digital cameras for use in airborne image acquisition has been favored in site specific irrigation management as it satisfies spatial, spectral and temporal requirements for precision agriculture. These cameras are low cost, and can provide real-time or near-real-time imagery. They can acquire imagery in very narrow spectral bands with high spatial resolution. Red, near-infrared and thermal images were taken from 2 different sites of cotton fields with variable irrigation treatments. Radiometric correction was carried out using ground calibration data. Data fusion technique was applied to develop a new indicator of crop water stress for irrigated cotton. It involved multispectral and multi-temporal data transformations dealing with crop temperature, vegetation ground cover, and soil brightness along soil line. The multi-dimensional fusion generates a spatial recognition of the degree of water stress, thus, being able to improve irrigation strategy.