Objective To address the path planning challenges for robots in the ground-pot areas of light-flavor Baijiu fermentation workshops.Methods A robotic path planning method, named C-EAFO-YBWC, is proposed. First, gamma correction and unsharp masking are used to enhance image quality, combined with the adaptive FAST threshold to improve ORB feature point extraction, ensuring the localization accuracy of the ORB-SLAM2 algorithm under complex lighting. Second, BiFPN and Coordinate Attention (CA) mechanisms are integrated into YOLOv10n and optimized with WIoU to detect fermentation pots and generate path points. Finally, by combining robot self-localization and ground-pot detection, the path points are transformed into the robot's coordinate system to guide its motion.Results In the four test sequences selected from the EuRoC dataset, the RMSE is reduced by 2.60%, 43.26%, 12.72%, and 30.10%, respectively. In tests using a ground-pot dataset, mAP@0.5 improves by 1.1 percentage points, while Params and FLOPs are reduced by 8.33% and 2.32%, respectively.Conclusion The proposed EAF_ORB-SLAM2 and YOLOv10n_BWC algorithms effectively ensure the validity of path planning.