Due to the high complexity and diversity of photographic images, the integrity of the watermark is relatively low when under attack. For the above problems, a digital watermark algorithm of photographic image based on image segmentation technology is proposed. The Arnold transform is used to strengthen image security, RGB turns to YCrCb to optimize color processing, and two layers of discrete wavelet transform to deeply analyze the multi-scale features of the image. Through morphological corrosion optimization and intelligent merging strategy, the image is pre-processed to generate a clearly defined set of image blocks. The subblock with the highest contrast is selected, and the low frequency subband of discrete wavelet transform combined with quantitative jitter modulation technology is used to embed the watermark information in the image secretly and reliably. In the extraction stage, it integrates image segmentation, Daisy transform precise positioning, BSP tree fine division, and Arnold inverse transformation accurately extracts and restores the image watermark. The experimental results show that the design method can significantly improve the robustness of watermark in multiple attack scenarios, and ensure the integrity and reliability of watermark information.