The infrared imaging technique can be used to image the temperature distribution of the body. It's hopeful to be applied to the diagnosis and prediction of many diseases. Image processing is necessary to enhance the original infrared images because of the blurring. In this paper, the image enhancement technique based on the Retinex theory is studied. The algorithms such as Frackle-McCann algorithm, McCann99 algorithm, single-scale Retinex algorithm and multi-scale Retinex algorithm are applied to the enhancement of gray infrared image. The acceptable results are obtained and compared.