Evaluation of imaging in automated visual web inspection: Dissertation

In this research methods for the evaluation of the imaging in automated visual web inspection are presented. In web inspection a flat object is moving in the field of view of the sensor in such a way that an image can be formed by synchronizing the image formation and the web movement. Examples of web-like objects are metal, paper, textile and plastic strips or sheets and wooden boards or planks. The research concentrates on the evaluation of a CCD or TDI line scan based imaging unit. An imaging unit will be treated as a system consisting of imaging optics, camera, digitization electronics and mechanical and other accessories. Several methods for the quantification of the properties of an imaging unit are introduced. These include a method for the determination of the system modulation transfer function and for the determination of the system noise characteristics. In addition, the determination of the response linearity and uniformity across the image area is discussed in the context of the system dynamic range. The properties of the exemplary systems are derived in most cases from digitized test image data. Light sources and illumination arrangements are considered only if they affect on the quality of imaging. It is assumed that light sources used in practical CCD or TDI line scan based inspection are incoherent and polychromatic. The main objective of the research is to demonstrate that the performance of an imaging unit of a web inspection system can be characterized with a few figures of merit which can be used for systematic evaluation and design of imaging in web inspection applications. Application-specific examples of the systematic evaluation of imaging in web inspection are given.