云纹
平行四边形
图层(电子)
几何变换
基础(拓扑)
转化(遗传学)
几何学
流离失所(心理学)
计算机科学
计算机视觉
光学
图像(数学)
数学
人工智能
材料科学
数学分析
物理
心理学
生物化学
化学
机器人
心理治疗师
复合材料
基因
标识
DOI:10.5075/epfl-thesis-3434
摘要
We explore a new approach for synthesizing moire images that can be used for authentication of documents. For synthesizing moire images, we need two layers: a base layer made of replicated bands or parallelograms and a revealing layer made of transparent lines (or dots). When superposing a base layer and a revealing layer, moire images appear. Our method enables us to create dynamically moving messages incorporating text or color elements. In a simple layout of the base layer, the same pattern is replicated in each base band or parallelogram. Since the base layer's replication periods and the revealing layer period are similar, the revealed image is formed by the replicated tiny patterns, enlarged and possibly transformed. By considering the formation of the moire image as a line sampling process, we derive the linear transformation between the base layer and the moire image. We obtain the geometric layout of the resulting moire image, i.e. its orientation, size and displacement direction when moving the revealing layer on top of the base layer. Interesting moire images can be synthesized by applying geometric transformations to both the base and revealing layers. We propose a mathematical model describing the geometric transformation that a moire image undergoes, when its base layer and its revealing layer are subject to different freely chosen non linear geometric transformations. We derive the geometric transformation to apply to the base and revealing layers, in order to obtain a desired moire image transformation. We also derive layer transformations which yield periodic moire images despite the fact that both the base and the revealing layers are curved. The approach for deriving the relationship between the respective geometric transformations of base layer, revealing layer and moire images, relies on the comparison between the relative shifts of the revealing layer, the base layer and the moire image. We generalize the 1D moire synthesis approach to the 2D moires. The base layer is made of replicated tiny parallelograms. The revealing layer is made of transparent dots. New interesting effects are obtained by applying geometric transformations such as independent moire movements in two different orientations. A further method is presented in chapter 5, which relies on elevation profile level lines. A set of lines is modified according to an elevation profile. When superposing the modified base layer and the revealing layer, the level lines are revealed by a succession of various intensity or color lines. This elevation profile can either be a mathematical function, or can be created from an image containing text or symbols. Most moire effects are created thanks to an interactive Photoshop plug-in called MoireImaging.
科研通智能强力驱动
Strongly Powered by AbleSci AI