Design of AGV Low-Delay Visual Servo System

The visual sensing technology has been more and more applied to AGV (Automated Guided Vehicle), sometimes, the visual-aided navigation will fail to orientate and grab the objective due to vehicle body shaking. For the traditional method, the motion characteristics between image frames are obtained through Fourier transform of images, to obtain the image compensations and correct the image, with more calculations and poor real-time performance. In most AGVs, gyroscope and pose measurement sensor are equipped. In this paper, the output parameters of gyroscope were read, to correct the image and obtain the stable output. Through experiment, time delay for obtaining the corrected parameters is reduced by 97% than that obtained through traditional method.