In the process of defect detection of mobile phone screens using traditional algorithms, small defects are easily overlooked and erroneously detected. Compared with other algorithms, the two-dimensional Otsu segmentation algorithm adds neighborhood threshold information, which reduces the influence of noise on the detection results, but the detection speed is slow. The author proposes an improved 2D Otsu fast defect segmentation algorithm. First, bilateral filter and wavelet denoising algorithms are used to remove more noise. Then, the combination of straight line intercept and Particle swarm optimization is used to improve the Otsu algorithm to quickly and accurately segment the target and background. Finally, the author conducted an experiment to apply the algorithm in this paper and the traditional Otsu algorithm to mobile phone screen detection, and compared the segmentation results. The results show that the improved algorithm in this paper optimizes the inspection accuracy, improves the segmentation efficiency, and reduces the running time.