Enhancing 3D Reconstructions in Underwater Environments: The Impact of Image Enhancement on Model Quality

Abstract. 3D reconstructions in underwater environments face significant challenges due to poor image quality, caused by blurring, reduced contrast, color distortion, and inadequate lighting. This study investigates the impact of various image enhancement techniques on underwater 3D reconstruction, focusing on Contrast Limited Adaptive Histogram Equalization (CLAHE), RGB Histogram Stretching (RGHS), and a combined approach integrating RGB stretching with CLAHE. Three real-world underwater datasets were analyzed to assess the effectiveness of these methods in improving the accuracy and completeness of reconstructed 3D models. Notably, the RGB-CLAHE combination achieved the most substantial improvements, increasing reconstructed points by 7.60%, detected features by 7.56%, and reconstructed features by 12.94% on average across the datasets. However, the enhancement methods did not improve performance in certain cases, underscoring the need for a robust evaluation methodology to determine the most suitable technique for each dataset. These findings support the design and development of adaptive evaluation frameworks alongside advanced enhancement strategies, enabling more accurate and detailed analysis of submerged environments. Such approaches have the potential to benefit a wide range of scientific and industrial applications, including marine research, underwater archaeology, and inspection tasks.