Abstract
This paper presents a distortion compensation algorithm for multi-camera systems that does not require specialized calibration targets. A review of classical distortion models as well as modern calibration approaches (such as OpenCV) is provided. A modification of the distortion compensation algorithm is proposed, and an experimental comparison with the classical Brown–Conrady model is conducted using data from cameras equipped with the Sony IMX462 sensor. The results show that the proposed algorithm achieves projection rectification comparable to target-based calibration while preserving a wide field of view. This opens up opportunities for applying the method in wearable monitoring systems, panoramic image creation, and pipeline inspection, where the use of calibration targets is challenging.