AuthorLandsiedel, Emma Catherine
AdvisorKoshel, Richard J.
MetadataShow full item record
PublisherThe University of Arizona.
RightsCopyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction, presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
AbstractAn extended depth of field imaging system was developed for in-line inspection for the semiconductor industry. The system produces a single, two-dimensional, in-focus image of objects such as integrated circuits or other computer chip components. The system must be high resolution and have a large depth of field to capture small details at all heights of the object. The system must operate quickly since it is used on an assembly line and also must produce true color images, since color can be an indicator of issues with an inspected object. The system uses white light interferometry (WLI) as a means of finding focus across the field of view. The interferometer vertically scans through different focus positions. Fringes from WLI only occur near best focus, or zero optical path difference (OPD). A “vision sensor,” which detects changes in irradiance, outputs events when the irradiance changes for each pixel, which only occurs in a narrow depth fringe region. A depth map of the object is created from this information and informs a separate imaging sensor of best focus depth for individual pixels. The imaging sensor captures pixels at their best focus, and the scanning combines the individual pixel information to create the in focus, two-dimensional image. This thesis details the development of the system concept, the design of the system, and modeling of white light interference in general as well as its use in modeling system performance. The system designed has a 0.6-mm x 0.8-mm field of view with 1.5-μm object size mapping to an image sensor pixel. The system has a 500-μm depth of focus with depth resolution of 3 μm and creates true color images. The prototype of the system is expected to create the final image in around 1 second. The speed is limited by capabilities of current detectors rather than the method itself, and it is expected the system can become significantly faster in a few years, such that full image capture can occur in a time of 40 ms.
Degree ProgramGraduate College