Tray identification method, pallet forking method and autonomous mobile forklift

The invention provides a tray identification method, a method for forking a tray by using a forklift, an autonomous mobile forklift, electronic equipment and a non-instantaneous computer readable storage medium. The tray identification method comprises the following steps: determining at least three visual key points of a tray; acquiring laser point sets respectively corresponding to the at least three visual key points in a laser point cloud; determining a laser representative point in each laser point set; and performing linear constraint filtering on the laser representative points in the laser point set, and determining the representative points at the two ends in the representative points subjected to linear constraint filtering as the two end points of the tray. 本申请提供了一种托盘的识别方法、利用叉车叉起托盘的方法、自主移动叉车、电子设备以及非瞬时性计算机可读存储介质。托盘的识别方法包括：确定托盘的至少三个视觉关键点；在激光点云中获取分别与所述至少三个视觉关键点相对应的激光点集；在每个所述激光点集中确定激光代表点；以及对所述激光点集中的激光代表点进行直线约束过滤，并将通过所述直线约束过滤的代表点中两端的代表点确定为所述托盘的两个端点。