This paper proposes a data cleaning method combining the Local Outlier Factor with Image Threshold Segmentation (LOF-ITSM), in order to address the problem that traditional image threshold segmentation methods cannot effectively detect outlier data in wind power curves, and global thresholds are difficult to adapt to curve fuzzy boundaries. Firstly, the LOF algorithm is used to detect wind power data and identify outlier types of abnormal data; Then, by introducing a local adaptive threshold optimization mechanism, the image threshold segmentation method is improved to better adapt to the fuzzy boundaries in the curve, while alleviating the sensitivity of the LOF algorithm to stacked abnormal data. The experimental results show that this method can better detect abnormal data in the wind power curve, laying a data foundation for promoting efficient utilization of wind energy and green transformation of energy structure.