Approach and Fork Insertion to Target Pallet Based on Image Measurement Method

Previously, we proposed a navigation method based on image processing to move to the front of a pallet that can be detected in the field of view. We also proposed an image-processing method to estimate the inclination angle when the pallet is inclined at a pitch with respect to the running surface, such as a pallet loaded on a truck. In this study, we improved the robustness of the existing method so that the series of operations from approach to fork insertion can be realized stably without being affected by the environment. Furthermore, a series of operations from approach to fork insertion was realized by an automatic forklift (Autonomous Guided Forklift, AGF) in an indoor laboratory space simulating a warehouse.