当印制在商品、证件和文档等物品上的光学全息图被拍摄或复制时，由于光学成像质量或拍摄条件的影响，采集的全息图像会产生畸变。据此，提出了一种全息图失真校正的方法。该方法首先运用图像处理技术，提取出所需要的全息图目标场景，然后在被提取出的全息图上确定全息图场景的边界，并在边界上自动搜索到4个顶点和4个中点的位置以及相对应的理想点的位置，最后进行投影失真校正和非线性失真校正。与其他图像畸变校正方法相比，该方法无需准备高精度标定模板，且校正和识别过程中只需用到一幅目标全息图像。实验证明，该方法能够很好地识别畸变的光学全息图，且具有效率高、快捷、识别效果好的优点。

Images captured from optical holograms printed on goods, documents, personal identity cards, etc. would be distorted due to the imaging quality of the optical system and the influence of shooting conditions. A distortion correction method for the captured optical holograms is proposed. Firstly, an image pre-processing technique is used to extract the target hologram and the boundary of the target hologram is determined. Then four vertexes and midpoint positions of the hologram and the corresponding ideal points can be sought automatically. The holograms with projection distortion and nonlinear distortion are corrected by using different polynomial models. Compared with the other methods, the proposed one is simpler and faster, as it does not need the high-precision calibrating module and deals with an object′s hologram only in calibrating and recognizing process. The experiments show that the corrected holograms can be recognized clearly. The proposed method has the advantages of high efficiency, fast speed and high-recognition rate.