| Summary: | 碩士 === 國立中央大學 === 資訊工程學系碩士在職專班 === 91 ===
Web-based pattern recognition (PR) is a novel method for multi-platform in real-time Internet. This study attempts to use associative memory to apply pattern recognition technology to the real-time pattern recognition in a web-based recognition system with a Client-Server network structure. Remote user can draw the shape of the engineering components or characters using the browser, and the recognition system then searches the database via the Internet. The server-end includes databases for storage of a sample pattern. With training, the user can assign any pattern to what in real-time. The pattern is recorded in the server-end databases. With respect to retrieval tasks, a novel PR method is proposed that depends on matching databases. The method can efficiently solve the problem of spurious states from recurrent neural network (RNN) in the Web-based PR (WBPR) system. Database matching overcomes the capacity restrictions on RNN. In the new approach, the WBPR system divides the set of records in the server-end databases, and determines W and □ from different segments. The WBPR system finally employs various segments to perform PR tasks. Two simulation experiments are also discussed to clarify and corroborate the above Web-based PR technology. The first experiment is some recognition of characters. The second experiment is implemented on the Yang-Fen Automation Electrical Engineering Company as a case study. The case is continued for four months, and engineers are also used to operating the web-based pattern recognition system. Therefore, the cooperative plan described above is analyzed and discussed here. Finally, these papers propose the tradition methods compare with the innovative methods.
|
|---|
