Search

A new architecture for statistical classification of hand-printed characters is presented. It is based on standard preprocessing and three feature types, containing geometrical information on the position of the pixels, on the contour orientation and on the bending points, respectively. Two feature vectors at a time are used as inputs to a multi-layer perceptron-based classifier, giving rise to three simple classifiers operating in parallel. The outputs of the three different classifiers are mixed by a final supervisor realized by a perceptron layer. The overall network has been trained using digits, upper and lower case letters of the NIST Special Database 3. Classification results of the NIST Test Data 1 are provided. The system has an error rate of 2.59% on the digits of NIST Test Data 1 at zero rejection rate, while it has 2.99 and 11.00% error rate on the upper and lower case letters, respectively.

A new statistical classifier for handwritten character recognition is presented. After a standard preprocessing phase for image binarization and normalization, a distance transform is applied to the normalized image, converting a black and white (B/W) into a gray scale picture. The latter is used as feature space for a k -Nearest-Neighbor classifier, based on a dissimilarity measure which generalizes the use of the distance transform itself. The classifier has been implemented on a massively-parallel processor, Connection Machine CM-2. Classification results of digits extracted from the U.S. Post Office ZIP code database and the upper-case letters of the NIST Test Data 1 are provided. The system has an accuracy of 96.73% on the digits and 94.51% on the upper-case letters when no rejection is allowed and an accuracy of 98.96%, on the digits and 98.72% on the upper-case letters at 1% error rate.

Electromagnetic interference may cause failures in operational amplifiers. The probability of these failures can be reduced by properly designing the opamp, once the failure mechanism has been discovered. In this paper the design of some integrated BiCMOS operational amplifiers with a very low-probability of electromagnetic interference (EMI) induced failures is reported. In particular, it is shown that opamps exhibiting good general performances as well as low EMI-susceptibility can be obtained only if their response to a large square-wave input signal is symmetric and the influence of some parasitic capacitances in the input stage is compensated. Following these guide-lines, we found possible to design BiCMOS opamp structures exhibiting EMI susceptibility of only a few mV up to several hundred MHz when they are driven with an interfering input signal of some volts.

Failures induced on analog integrated circuits by electromagnetic interference (EMI) will be analyzed with particular emphasis on integrated operational amplifiers built with different technologies. Additionally, the correlation found between EMI susceptibility and large-signal opamp behavior will be discussed. Some criteria for the design of low EMI susceptibility opamps will be derived. Finally, as an application example, the design of a BiCMOS opamp with an extremely low-probability EMI-induced failure will be presented.

Handwritten digit and alphabetic character recognition has been an active area of research for many years. Character recognition techniques associate a symbolic identity with the image of a character. The problem of replication of human functions by machines involves the recognition of both machine and hand-printed or cursive-written characters. Character recognition is better known as Optical Character Recognition (OCR) since it deals with recognition of optically processed characters rather than, for example, magnetically processed ones. The origin of character recognition can be found as early as in 1870, though the modern version of OCR appeared in the middle of the 1940s with the development of digital computers. Its main area of interest was data processing with application to the business world. The principal motivation for development of OCR system is the need to cope with the enormous flood of paper such as bank cheques, commercial forms, government records, mail sorting generated by the expanding society and international partnerships. Since the origin of OCR systems extensive research has been carried out and a large number of technical papers have been published by researchers in this area. Many techniques and experimental results have been published in the literature. In recent years, new classification approaches, such as neural networks, and design methodologies, such as classifier combination and parallel feature networks extraction, have renewed interests in this fields. As OCR technology seeks applications in more complicated, real time environments, the criterion for successful algorithm becomes more rigorous. New feature extraction and classification techniques are developed with parallel hardware implementation in mind. Several recognition methodologies are combined in order to improve the recognition quality. State of the art reports on character recognition research have been presented by Nagy [1968], Harmon [1972], Suen et al. [1980], Mori et al. [1984], Mantas [1986], Davis and Yall [1986], Govindan and Shivaprasad [1990], Lee and Srihari [1993].