Visual cryptography is an encryption technique that decomposes secret images into multiple shares. These shares are digitally or physically overlapped to recover the original image, negating the need for complex mathematical operations or additional hardware. There have been many variations of visual cryptography proposed over the years, each addressing different problems or to fulfill different security requirements. Existing review papers on the area only cover certain types of visual cryptography or lack comparisons between the various schemes. To address this gap, this paper provides broad overview of the area to aid new researchers in identifying research problems or to select suitable visual cryptography methods for their desired applications. For more veteran researchers in the area, our paper provides the most up-to-date coverage of the state-of-the-art. We first provide an introduction to the various categories of visual cryptography techniques, including a discussion on recently proposed schemes. These schemes are then compared in terms of their features, performance metrics, advantages and disadvantages. Compared to prior work, we extend the number of comparison metrics to include signal-to-noise ratio and the type of shares. Over 40 visual cryptography schemes that have been proposed in the past two decades were analyzed and compared. Our findings indicate that existing problems such as pixel expansion, poor quality of recovered image quality, computational and memory complexities still exist, and a optimizing the trade-off between these requirements still requires further investigation. We conclude the paper with a discussion of these open problems and future research directions.

视觉密码学是一种将秘​​密图像分解为多个共享的加密技术。这些共享在数字或物理上重叠以恢复原始图像，不需要复杂的数学运算或额外的硬件。多年来，已经提出了许多视觉密码学的变体，每个变体都解决了不同的问题或满足不同的安全要求。该领域现有的评论论文仅涵盖某些类型的视觉密码学或缺乏各种方案之间的比较。为了解决这一差距，本文提供了该领域的广泛概述，以帮助新研究人员识别研究问题或为其所需应用选择合适的视觉密码学方法。对于该领域的更多资深研究人员，我们的论文提供了最新技术的最新报道。我们首先介绍各​​种类别的视觉密码技术，包括对最近提出的方案的讨论。然后比较这些方案的特性、性能指标、优缺点。与之前的工作相比，我们扩展了比较指标的数量，以包括信噪比和共享类型。对过去 20 年提出的 40 多种视觉密码方案进行了分析和比较。我们的研究结果表明，像素扩展、恢复图像质量差、计算和内存复杂性等现有问题仍然存在，优化这些需求之间的权衡仍需要进一步研究。