情報処理学会 第76回全国大会講演要旨
4Q-5
顔と位置情報に基づく写真の自動分類と画像検索
Nowadays the number of people who have cameras, either digital one or one included on the mobile phone, is increasing. People document their life by taking a lot of photos per day. Storing and organizing such a huge amount of photos can be a challenging task. It will be hard to retrieve a desired photo taken in a certain place with a certain person. Most of the current available image classification systems use manual annotation to tag the GPS data though it’s a time consuming task. Thus, we developed a system which can organize photos by categorizing to faces and places.
For place detection, GPS data is used. In case photos contain no GPS data, image-matching techniques find similar geo-tagged photos in the database to the user photo to retrieve GPS data. To detect and identify faces in the photo, similar image-matching technique are used.
For place detection, GPS data is used. In case photos contain GPS data, GPS data is extracted from the photos. In case of no GPS data, image-matching techniques find similar geo-tagged photos in the database to the user photo to retrieve GPS data.
In this paper, mainly we talk about image based retrieval techniques used in our system.
For place detection, GPS data is used. In case photos contain no GPS data, image-matching techniques find similar geo-tagged photos in the database to the user photo to retrieve GPS data. To detect and identify faces in the photo, similar image-matching technique are used.
For place detection, GPS data is used. In case photos contain GPS data, GPS data is extracted from the photos. In case of no GPS data, image-matching techniques find similar geo-tagged photos in the database to the user photo to retrieve GPS data.
In this paper, mainly we talk about image based retrieval techniques used in our system.
