E-Thesis & Research สถาบันเทคโนโลยีไทย-ญี่ปุ่น

fieldid E-Thesis & Research สถาบันเทคโนโลยีไทย-ญี่ปุ่น

เขตข้อมูล	ข้อมูล
บทคัดย่อ	SEMI-AUTOMATIC KNOWLEDGE REPRESENTATION FOR SEMANTIC-BASED IMAGE RETRIEVAL EMPLOYING TOPIC MAPS TECHNOLOGY : Many pieces of research work on image retrieval have been extensively conducted in past decades employing various methods including simple keyword tagging and content-based image retrieval. However, some gaps are still presented as those mentioned methods are missing semantic approach. Basically image search is quite a complex area as image elements require sophisticated algorithm to extract, analyze and find ones with a closest match. The semantic-based Topic Maps are used to identify the image elements to further understand the image content and relationship between each image composition. Thus, this research has been done to prove the concept by implementing Topic Maps related processes on top of a conventional content-based image retrieval engine. Experimental results show that the proposed semantic-based image retrieval system is able to improve the image interpretation. With this way, the proposed image retrieval system is able to process image low-level elements while maintain high-level understanding to produce better result of precision and recall.
ผู้แต่ง	Siriporn Chongpratheep
ประเภทสิ่งพิมพ์	วิทยานิพนธ์ (MIT) 2015
ปีที่พิมพ์	1
เลขหน้า	117 p.
หัวเรื่อง	Industrial Management--Thesis
หัวเรื่อง	Thai-Nichi Institute of Technology.Industrial Management--Thesis
เอกสารฉบับเต็ม	Download full text

วิทยานิพนธ์

สารนิพนธ์

สหกิจศึกษา

งานวิจัย

บทความวิจัย