This paper presents the illumination design of a real-time live-bird imaging system for determining the size and initial presentation of a bird on a moving conveyor. A real-time live-bird imaging system presents a challenging design problem, for it must minimize the variability of the birds’ visual reflexes to mechanical processes, it must account for variations in bird size/shape/color, it must meet the cycle-time requirement, and yet provide an adequately illuminated environment to ease human supervision. In this paper, we first identify the variables needed for motion prediction. Second, by analyzing the bird visual perception we have developed a two-stage structured illumination that has the potential to minimize the demand on the control efforts of the transfer system, and to improve birds’ welfare and the ultimate product quality. Finally, we present the image algorithms and experimental results of the design evaluation using live birds. It is expected that the design principles presented in this paper provide essential bases for motion analysis, prediction, and control of an automated live-bird transfer process.