Optical Illusions Images Dataset

光学幻觉图像数据集由路易斯维尔大学创建，包含6725张幻觉图像，主要从两个网站收集而来。数据集旨在研究人类视觉系统对幻觉的感知，特别是颜色、大小、对齐和运动的误判。创建过程中，使用了网络爬虫技术从多个网站抓取图像，并进行了分类和元数据整理。该数据集应用于计算机视觉领域，旨在通过分析幻觉图像，提升计算机对视觉幻觉的理解和生成能力，从而推动计算机视觉技术的发展。

The Optical Illusion Image Dataset, created by the University of Louisville, contains 6,725 illusion images primarily collected from two websites. This dataset is designed to study the human visual system's perception of visual illusions, particularly misjudgments of color, size, alignment and motion. During its creation, web crawling technology was used to scrape images from multiple websites, followed by classification and metadata organization. Applied in the field of computer vision, this dataset aims to improve computers' ability to understand and generate visual illusions through the analysis of such images, thereby advancing the development of computer vision technology.