Stained Glass Collage

Automatic photo collages based on regions of interest

Stained Glass Collage uses regions of interest to automatically generate an attractive collage from a selection of photos.

We originally developed stained glass visualizations as a means to summarize videos by combining regions of interest in video shots into a collage. We adapted the video-based regions of interest algorithm to find regions of interest in any kind of photo after previously using face detection to find those regions in photos with faces. The revised algorithm uses a combination of texture and color to find regions of interest even in photos without clearly detectable faces.

As users may choose the orientation and aspect ratio of a collage, we need to place the photo regions on the canvas such that the available space is covered. To fill the spaces between regions, we assign each point on the canvas to the closest photo region for which the point is inside the photo bounds. To measure the distance between a photo region and a point in the canvas, we compute the Euclidian distance between the point and the center of the photo region and subtract the radius of the ellipse enclosing the photo region. This distance measure gives larger photo regions more coverage and creates slightly curved borders between areas. This process is illustrated the figure below. A point with the same distances to the three photo regions d1, d2, and d3 lies on the border of those photo regions.

After an initial collage is generated, users may change the orientation and aspect ratio of the collage and select border thickness and color. We provide a Java applet that lets users reorder and remove images through drag-and-drop operation. Users can also change the region-of-interest by moving the handles of a bounding box. Finally, users may zoom in or out of individual images.

Technical Contact

Related Publications

Publication Details
  • UIST 2004 Companion, pp. 13-14
  • Oct 24, 2004


We developed a novel technique for creating visually pleasing collages from photo regions. The technique is called "stained glass" because the resulting collage with irregular shapes is reminiscent of a stained glass window. The collages reuse photos in novel ways to present photos with faces that can be printed, included in Web pages, or shared via email. The poster describes the requirements for creating stained glass visualizations from photos of faces, our approach for creating face stained glass, and techniques used to improve the aesthetics and flexibility of the stained glass generation. Early user feedback with face stained glass have been very positive.
Publication Details
  • Proceedings of 2004 IEEE International Conference on Multimedia and Expo (ICME 2004)
  • Jun 27, 2004


This paper presents a method for creating highly condensed video summaries called Stained-Glass visualizations. These are especially suitable for small displays on mobile devices. A morphological grouping technique is described for finding 3D regions of high activity or motion from a video embedded in x-y-t space. These regions determine areas in the keyframes, which can be subsumed in a more general geometric framework of germs and supports: germs are the areas of interest, and supports give the context. Algorithms for packing and laying out the germs are provided. Gaps between the germs are filled using a Voronoi-based method. Irregular shapes emerge, and the result looks like stained glass.