Much of our research focuses on image enhancement based on image content. By detecting the presence of people, for example, one can improve how they look by automatically adjusting the color balance and contrast of their skin tones.
Once an image is captured, this type of analysis has further applications. Knowing where the people are in a picture, for example, enables auto-cropping of an image to make a better-looking picture.
Image analysis also contributes to HP's media management solutions, which help people sort through their image libraries and automatically create engaging slideshows or other presentations. (See related information on photo and video collection management research.)
Another of interest involves turning raw sensor data into the best possible image inside a still or video camera. Similar problems need to be addressed in printers, where processing is required to convert a digital picture to printer dots.
In addition to these efforts, we are working with a group of technology, nonprofit and governmental organizations from around the world to build a permanent archive of multilingual digitized text and multimedia content.
As part of that group -- the Open Content Alliance -- HP Labs is contributing technologies to process scanned books and quickly create enhanced, compressed PDF files . |