Indexing Flowers by Color Names Using Domain Knowledge-Driven Segmentation.
MASSACHUSETTS UNIV AMHERST
Pagination or Media Count:
We describe a solution to the problem of indexing image of flowers for searching a flower patents database by ers iii t color. We use a natural language color classification derived from the ISCC-NBS color system and the X Win- this ap dow color names to effectively use the domain knowledge from tb available, provide perceptually correct retrieval and allow color of natural language queries. We have developed an autoception matic iterative segmentation algorithm with knowledge natnral driven feedback to isolate a flower region from the background. The color of the flower is defined by the color names present in the flower region and their relative proportions. The database can be queried by example and by color names. We demonstrate the effectiveness of the strategy on a test database.
- Information Science