Detail Evaluating cheese shred morphology and integrity using computer vision Bibliografi Author: Ni, Hongxu ; Gunasekaran, Sundaram (Advisor) Topik: AGRICULTURE; FOOD SCIENCE AND TECHNOLOGY|ENGINEERING; AGRICULTURAL (0539); COMPUTER SCIENCE Bahasa: (EN )    ISBN: 0-599-78245-5     Penerbit: THE UNIVERSITY OF WISCONSIN - MADISON     Tahun Terbit: 2000     Jenis: Theses - Dissertation Fulltext: 9972887.pdf (0.0B; 0 download) Abstract The algorithms for automatic quality inspection of cheese shreds are described. Two segmentation methods are proposed. In the thinning method, morphological elements are sequentially applied over the image to generate image skeletons. The occluded cheese shreds can be recovered by tracking the generated skeletons. However, our experiments showed that the thinning method is sensitive to the boundary noise. A more robust and efficient algorithm called the X-Y sweep method is proposed for better segmentation. In this method, the visual scene is swept in the x and y direction and two sets of run length code are generated. According to a width condition and spatial relations with the neighbor run length code, the run length codes are grouped as a block or a joint. A block is a segment with orientation preference. A joint is formed by collecting the pixels that can not be swept through in neither x nor y direction. The occluded cheese shred can be recovered by merging the neighbor blocks based on the local, semi local and global descriptions. The topological sorting method was used to find the best match. The algorithm was implemented with graphic interface support using Microsoft Foundation Class Library. The tests showed that the X-Y sweep method can correctly identify the cheese shred and similarly shaped objects. The tests with cheese shreds showed that the measurement error was within 5%. The design of a neural network model for associating cheese quality parameters with manufacturing conditions is also introduced. Opini Anda Klik untuk menuliskan opini Anda tentang koleksi ini! Lihat Sejarah Pengadaan  Konversi Metadata   Kembali

Copyright © 2006, 2007 Unika Atma Jaya, all rights reserved