Image analysis – Editing – error checking – or correction
Reexamination Certificate
1999-06-16
2003-02-25
Couso, Jose L. (Department: 2721)
Image analysis
Editing, error checking, or correction
C358S407000, C358S434000, C358S436000, C382S317000, C382S321000, C707S793000, C707S793000
Reexamination Certificate
active
06526184
ABSTRACT:
BACKGROUND OF THE INVENTION
The invention comprises a terminal which in the manner of the fax apparatuses which are conventional nowadays, reads in items of image information, effects conversion into codes by way of an OCR-unit, displays the result of that conversion immediately to the user of the terminal and thereby permits him upon detecting incorrect conversion of his items of text information into codes either to repeat the input operation or however also to send an input which is not correct to the target address in such a way that he informs the receiver about the defectiveness of the conversion in order to permit manual post-processing.
Unlike the input of commands or items of information on computer installations by making use of an alphanumeric keyboard and a display screen, which is still the predominant procedure nowadays, such inputs can also be implemented by way of written, more specifically manuscript entries, on paper. The software for effecting the process for converting writing into codes is summed up by the term OCR (optical character recognition). It identifies a well-documented discipline which includes processes and procedures, methods (feature recognition, pattern recognition), error-tolerant interpolation processes (fuzzy logic) and technical aspects of data compression, optimisation of implementation times etc.
The technology of OCR however encounters natural limits which exclude completely error-free results; on the one hand human handwriting has script images which are individual to a person and on the other hand it is not always unchanged even in relation to the same person, at different times.
This represents a serious obstacle in regard to the use of manuscript or handwriting as an input instrument; even very low error rates which will never be entirely eliminated hitherto made it impossible to put such terminals to practical use. Otherwise in many cases, in particular in the area relating to simple everyday uses as occur for example in the self-service sector in banks, such terminals would already long ago have made greater inroads into use of the equipment of keyboard and display screen. The psychological advantage of handwriting input is also that the input medium, that is to say the form, remains in the hands of the user in its original as a physical evidence and documentation of that which was communicated to the input unit as a command.
In order to illustrate the hitherto state of the art, insofar as it is to be deemed useful for comprehension of the present invention, a brief outline of the development of input peripheral units will be set forth. The Seventies were in the sign of the large computers to which inputs or commands were communicated by way of ‘unintelligent’ terminals. In the Eighties the Personal Computer began its triumphal march, a piece of equipment of high decentral intelligence, which was also very quickly used as a decentral peripheral unit and extensively displaced the other terminals.
It is precisely in regard to that function however which increasingly gained in significance that the weaknesses and problems of the PC manifested themselves. The PC is basically both too complicated and also too expensive just for communicating simple inputs and commands to central computers (hosts and servers). In that respect, not only equipment supply costs but more also maintenance costs and also the level of obsolescence which is high due to fast system and product cycles play their part. As from about the year 1995 such realisations led to the development of ‘lean clients’. The expectation is that they will win through, for cost reasons, particularly in relation to simple uses. More specifically, users are either not always in a position or not always prepared to satisfy high demands in terms of operating know-how. In regard to everyday processes in the self-service sector and also information access procedures, for example over the Internet, in the long term simple processes involving easy comprehensibility, simple handling and rapid implementation will win through. Even users who are accustomed to more complicated operating procedures, for example on a PC, will prefer the simpler technology if the straightforward use of the application is the foreground consideration and not the play component of operation which is also inherent in the PC, if therefore there is a realistic weighing-up of cost and use.
The experiences and realizations of recent years have resulted in two evolutionary developments. First, the symbol-oriented graphic user interface together with an expansion of the keyboard, by virtue of mouse operation, has made computer technology easier and more user friendly. Second, the development of personal computers (PC's) that provide a simpler and less expensive alternative to ever increasing high-capacity centralized computers, has allowed computer technology to promulgate.
These experiences and realisations resulted in recent years in two evolutionary developments. On the one hand, the symbol-oriented graphic user interface together with an expansion of the keyboard by virtue of mouse operation, in the meantime the virtually monopoly domain of Microsoft. Secondly, the development of so-called ‘lean clients’ as a simpler and in particular less expensive alternative to ever increasingly high-capacity decentral computers, that is to say PCs.
Today, in the strategically crucial area of information technology, the man-machine interface, we are faced with revolutionary developments, more specifically communication of man with machine by way of speech and also by way of handwriting. On a world-wide basis, great endeavours are being undertaken to achieve advances in this area—nonetheless an interface design which is as user-close as possible will afford a quantum leap in many application areas.
Speech and handwriting input can be viewed as equivalent processes, even if each has its own different advantages and disadvantages and therefore application areas which are rather complementary. In regard to preferences for the two methods, there is a marked difference: while in European and American research and development, the talk is exclusively of ‘speech input’, the Japanese electrical and electronic industry is targeted predominantly on ‘writing input technology’. That reflects cultural differences. While in the West a high level of affinity for abstract thought prevails, which devotes itself to virtual space, Japanese thinking is more concrete and practice-oriented and more remote from theory. A significant economic consequence thereof was the fact that the fax apparatus was developed in Europe but taken up in Japan and from there was made into a world-wide market success.
Virtually all known patents in the field of writing processing are Japanese, more specifically for many years now, from the introduction of the fax apparatus. It may be that this Japanese dominance, this advance in development, was also a reason why Western research fell back into the area of speech and ‘writing recognition’ is not a theme that is involved there. The converse is not the case. There is greatly advanced Japanese research and development in the field of ‘speech recognition’, and for that reason it must also be reckoned that this technology will first be applied in Japan; from the point of view of its structure and its auditory aspects the Japanese language is many times easier to detect by machine than Western languages.
The most recent trend in development towards ‘lean clients’ in combination with handwriting process has directed Japanese research and development to a technology and equipment which would open up attractive possible uses; the known fax unit, unused in its potential function as a ‘lean’ input peripheral unit, transports writing and thus potential control commands or items of input information. In that case the writing or text can be converted into codes either at the transmission location or at the target location, by means of OCR. If the conversion operation is moved into the target system, it would be possible for all pre
Altera Law Group LLC
Couso Jose L.
Desire Gregory
LandOfFree
Error-tolerant image computer terminal does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Error-tolerant image computer terminal, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Error-tolerant image computer terminal will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3166412