Abstract: In order to solve the problems of colour mixture, arc transition and multi-colour fine spraying in the camouflage spraying process by spray-painting robot, a rectangular gun with right-angle boundaries is designed to prevent overspray, and its coating deposition model is established. The performance test of the spray gun shows that the coating thickness of the improved rectangular gun is uniform, which meets requirements of the spraying process of the digital camouflage pattern. Secondly, the corner coordinates of the digital camouflage pattern are extracted by the corner detection algorithm. Meanwhile, the grid map of the digital camouflage pattern is established based on the corner coordinates. Afterwards, an improved BINN (biologically inspired neural network) is proposed. The propagation direction of the neural activity is restricted to 4 orthogonal directions and the range of the dynamic activity landscape is reduced by the robot movement rule combining with the spray-painting process of digital camouflage, which reduces the computational complexity. Finally, the two-dimensional path is mapped to the surface of the vehicle model to generate the three-dimensional camouflage spraying path based on the matrix transformation method. The effectiveness of the proposed method is verified by simulation and digital camouflage spraying experiment, the computational complexity of the BINN is decreased, and the problems of overspray and spraying path planning of the digital camouflage in robotic spraying are solved.