This instrumentation award is for equipment to start a laboratory in computer vision and image processing. The research projects include projects on face recognition, image algebra, and digital image processing for end-to-end systems. The face recognition problem involves studying issues in detection, identification, classification, and analysis of human faces. The image algebra research aims to develop the necessary efficient implementations of algebraic operations necessary to perform common image processing tasks such as magnification, shrinking, and connected component labeling. The digital processing in end-to-end systems consists of the necessary algorithms to obtain high quality images in the presence of noise form components such as lens and transmission lines. Research in vision encompasses a broad array of tasks. This research project addresses three such tasks; face recognition, image algebra, and digital image processing in end-to-end systems. Face recognition is an important problem with security applications. It is a task that humans are particularly good at and yet has proven very hard to computerize. Work in image algebra is useful in defining operations common to visual processing tasks. Such tasks include the magnification, rotation, and labeling of image regions. The implementations of these operations should be as efficient as possible. The processing of images using digital techniques involves the use of many devices that may introduce noise or be subject to measurement errors. These include the device that measures the intensity and color of the image, any lenses involved in the capture of the image, and transmission lines. Each of these can be treated separately, but the proper treatment is to consider the system as a whole as a means to obtain the highest quality images.
