Infrared imaging technology is widely used in many fields, but the non-uniformity of the design and manufacturing process seriously reduces the imaging quality of infrared images. To solve this problem, an infrared image correction system based on FPGA is designed. The system uses a two-point correction algorithm to update the pixel gain and bias of a domestic short-wave detector, then uses the 3σ algorithm to detect blind pixels and uses the idea of domain substitution to correct the blind pixels. The blind pixels are replaced by the values of valid pixels around the blind pixels in proportion to reduce the operational complexity. Experiments are carried out on the hardware system based on XC5VLX110T FPGA. The results show that the designed infrared image correction system can improve the image non-uniformity, the image quality and the imaging speed. In addition, the abundant spare resources make the system have better expansibility. This system has certain application value in infrared image processing.