Abstract: To obtain the boundary range of cornfields in large fields in an efficient and automated manner, this study used unmanned aerial vehicle (UAV) light detection and ranging (LiDAR) to obtain high-density point clouds. After point cloud filtering and elevation normalization, the maximum inter-class variance (Otsu) algorithm was improved and applied in combination with morphological opening and closing operations and Canny operator for detection. A cornfield boundary recognition method based on UAV LiDAR and improved Otsu algorithm was proposed, with experimental validation conducted in a selected orchard study area. The results showed that the combination of UAV LiDAR and the improved Otsu algorithm could iteratively determine the optimal threshold for cornfield recognition. After edge detection, the cornfield boundaries in the field environment could be accurately recognized. The proposed method was validated by sampling orthophoto images in the experimental area, and the results confirmed its high accuracy in cornfield boundary recognition, verifying the method's effectiveness. This study offers valuable references for engineering applications in cornfield boundary recognition, crop yield estimation, and smart agriculture research.