Facial recognition system

The pioneers of automated face recognition include Woody Bledsoe, Helen Chan Wolf and Charles Bisson.

During 1964 and 1965, Bledsoe, along with Helen Chan and Charles Bisson worked on using the computer to recognize faces (Bledsoe 1966a, 1966b; Bledsoe and Chan 1965). He was proud of this work, but because the funding was provided by an unnamed intelligence agency that does not allow a lot of publicity, was a bit of work published. Given a large database of images (actually a book of mug shots) and a photograph, the problem was to select from the database a small set of items so that the image records matched the photograph. The success of the method can be measured in terms of the relationship between the response list to the number of entries in the database. Bledsoe (1966a) discloses the following difficulties: 2

This project was labeled man-machine because the human extracted the coordinates of a set of features of the images, which are then used by the computer for recognition. Using a drawing tablet (GRAFACON or RAND tablet), the operator will pull out the coordinates for features such as the center for the students, the inside corner, outside corner of the eye point of widows peak, and so on. From these coordinates, a list of 20 intervals, such as the width of the mouth and eyes width, pupil to pupil calculated. These operators can process about 40 frames per hour. When you build the database name of the person in the image associated with the list of computed distances and stored in the computer. In the recognition phase, was seen at intervals in relation to the corresponding distance for each frame, yielding a distance between the photograph and the database entry. The next few items are returned.

Since it is unlikely that two images would match in rotation, tilt, tilt, and range (distance from the camera), each set of distances are normalized to represent the face in a frontal direction. To achieve this normalization program tries first determine inclination, lean, and rotation. Then, using these angles, the computer loosening effect of these transformations on the computed distances. To calculate these angles, the computer must know the three-dimensional geometry of the head. Because the actual heads were unavailable, Bledsoe (1964) used a standard head derived from measurements on seven heads.

After Bledsoe left PRI in 1966, this work continued at the Stanford Research Institute, primarily by Peter Hart. In experiments conducted on a database of over 2000 images, computer consistently outperformed humans when presented with the same information recognition (Bledsoe 1968). Peter Hart (1996) enthusiastically recalled the project with the exclamation: “”It worked really””

By about 1997 system developed by Christoph von der Malsburg and graduate students at the University of Bochum in Germany and the University of Southern California in the United States better than most systems with those of the Massachusetts Institute of Technology and the University of Maryland ranked next. The Bochum system was developed through funding by the US Army Research Laboratory. The software was sold as ZN-Face and used by customers such as Deutsche Bank and operators of airports and other busy places. The software was “”robust enough to make identifications from less than perfect view face. It can also often see through such barriers for identification as mustaches, beards, changing hairstyles and glasses with sunglasses””. [4]

In 2006, the results of the latest face recognition algorithms are evaluated Face Recognition Grand Challenge (FRGC). High resolution face images, 3-D face scans, and iris images used in the tests. The results showed that the new algorithm is 10 times more accurate than facial recognition algorithms in 2002 and 100 times more accurate than in 1995. Some of the algorithms could outperform human participants to recognize faces and identify the identical twins. [5] [6]

Related posts