Effective optical detection and imaging in challenging environments, such as turbid water and dense fog, remain a significant hurdle for conventional sensing systems. Light scattering and absorption severely degrade image quality and attenuate signals, compromising the performance of applications in autonomous navigation, underwater communication, and surveillance. Integral imaging (InIm) and lensless imaging—with advanced deep learning algorithms are two optical methods for addressing these issues. This review synthesizes the progression of this research, which showcases a clear trajectory from robust but computationally intensive 3D systems to highly efficient, compact, and noise-resilient lensless approaches for object and signal detection.