Viewing points of multi sub-images on display in distributed multimedia databases

Several video data contents retrieved from the distributed video databases are preferred to effectively be displayed as intended by the user. A screen composed of several sub-images is recommended to satisfy this requirement. An experiment was carried out to see how many video data contents could visually be recognized among sub-images at a time on a picture frame of the display, including their synopses. While subjects were viewing sub-images on the display, the distribution of the viewing points was measured. After the viewing tests, questionnaires of whether the subjects could correctly understand the contents of each sub-image were collected. Accordingly to the questionnaires, a picture frame consisting of four sub-images was found to be optimum.