| Summary: | 碩士 === 國立宜蘭大學 === 資訊工程研究所碩士班 === 99 === There are a lot of kinds of fishes in the aquarium. When you prepare to buy fish in the aquarium, you always want to get the fish’s information firstly. But you usually have no idea to describe the fish. Using the keyword ‘fish’ or ‘aquarium’ will get more result that we don’t need. The system offers content-based image search and retrieval.
In this paper, the system which takes picture as input the image of ornamental fish. Segmenting the region do Tamura texture feature, Auto-correlation, color feature and shape feature extraction. Analysis of four features results on a database of 410 fish images to 44 different kinds. The system has content-based image retrieval for the ornamental fish. It different searching process with traditional images retrieval of ornamental fish. The traditional images retrieval offer the several profile, user choose the closest. Then user selections the fish belong of subjects. Query to find relevant information.it difficult to the condition set with no relevant professional knowledge. The system can solve this problem. Just enter a picture of ornamental that the system can query target information. Even if user doesn’t need have professional, user can use the system quick search information about fish.
|
|---|
