Abstract: Infrared camera arrays utilize synthetic aperture technology to significantly enhance their ability to detect weak spatial targets, which has significant application value for the identification of weak and small infrared targets. However, the key to realizing multi-view synthetic aperture technology lies in obtaining high-precision camera calibration parameters. Currently, the most widely used camera calibration techniques suffer from insufficient accuracy, which limits the application range of infrared camera arrays. To address this issue, this study investigated an improved sparrow search algorithm based on tent chaotic mapping, along with a differential variation strategy to optimize the internal parameters of infrared camera arrays. The method first used the camera calibration method of Zhang to obtain the initial values of the internal parameters of the infrared camera array. The average reprojection error was then minimized using the improved sparrow search algorithm to further optimize the internal parameters. The experimental results indicated that applying tent chaotic mapping to population initialization, combined with a differential variation strategy, could significantly improve the search efficiency of the sparrow search algorithm, enabling it to maintain a global search capability while performing a local fine search when optimizing the internal parameters, thereby accelerating the convergence. Furthermore, compared with traditional calibration methods, the optimized internal parameters exhibited higher precision and better repeatability.