Rank-One Prior: Real-Time Scene Recovery

Scene recovery is a fundamental imaging task with several practical applications, including video surveillance and autonomous vehicles, etc. In this article, we provide a new real-time scene recovery framework to restore degraded images under different weather/imaging conditions, such as underwater, sand dust and haze. A degraded image can actually be seen as a superimposition of a clear image with the same color imaging environment (underwater, sand or haze, etc.). Mathematically, we can introduce a rank-one matrix to characterize this phenomenon, i.e., rank-one prior (ROP). Using the prior, a direct method with the complexity $O(N)$ is derived for real-time recovery. For general cases, we develop ROP$^+$ to further improve the recovery performance. Comprehensive experiments of the scene recovery illustrate that our method outperforms competitively several state-of-the-art imaging methods in terms of efficiency and robustness.

1 Introduction

An image is worth a thousand words. However, an image or a video captured in a turbid medium provides insufficient and incorrect visual information. For instance, Fig. 1 shows five images captured in different turbid media, such as sand dust, underwater, and haze. These images suffering from severe contrast and color alteration or degradation increases the difficulty in many important computer vision tasks, including image classification [1], object detection [2], surveillance systems [3], and semantic segmentation [4]. Hence, correct scene recovery from degraded observation is an essential and fundamental step in computer vision.