Image Retrieval by Using Shape Context
碩士 === 國立成功大學 === 電機工程學系碩博士班 === 94 === In this work we use shape context as our shape descriptor. The representation for a shape is a discrete set of n points. For each of these points, the shape context is a histogram of the relative positions of the remaining points. When a shape is rotated, th...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2006
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/01865808976323788027 |
| Summary: | 碩士 === 國立成功大學 === 電機工程學系碩博士班 === 94 === In this work we use shape context as our shape descriptor. The representation for a shape is a discrete set of n points. For each of these points, the shape context is a histogram of the relative positions of the remaining points. When a shape is rotated, the shape context is rotated too. We group the rotated shape contexts together and then label each group by an integer. Therefore, a shape is represented by a set of label. Using the histogram of label frequencies can quickly and efficiently search for similar or rotational shapes.
We use this shape retrieval method to integrate with an existent retrieval system which utilizes relevance feedback in region-based image retrieval. The system will learn the user’s semantic subjectivity. Hence, well accuracy is demonstrated in the results of image retrieval.
|
|---|