Original infrared images have a low contrast and high dynamic range, with the dynamic range varying between scenes. To obtain better images, it is necessary to transform the original infrared image data. Aiming to overcome the deficiency of the global histogram equalization algorithm, this study proposes a histogram equalization method based on region segmentation. By dividing the image into several regions, each region is histogram balanced; subsequently, pixels are reconstructed by linear interpolation. This effectively improves the image quality of the infrared thermal imaging system, greatly enhancing the contrast and image detail information. The algorithm is simple to implement in an FPGA and positive results are obtained.