Registers – Coded record sensors – Particular sensor structure
Reexamination Certificate
1998-07-31
2001-10-23
Frech, Karl D. (Department: 2876)
Registers
Coded record sensors
Particular sensor structure
C235S462160
Reexamination Certificate
active
06305606
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a process for scanning bar codes, and more particularly to a process for decoding a “deteriorated” bar code.
As is known, bar codes are widely used in the automatic identification of products.
A bar code is made up by a set of clear and dark zones, referred to in the following as space and/or bar type elements. Such elements have a rectangular shape and each of them may have a different width value.
A set of elements represents a code character univocally and a set of characters represents a symbol of a particular code univocally.
In fact, each code is defined by a set of rules that are the coding and decoding protocol for such particular type of code. Hence, a given set of characters that satisfies such protocol represents the decoded symbol of the code being examined.
Different processes for scanning and decoding bar code symbols are known in the art. A typical process may be summarised as follows: a bar code reading device illuminates the code to be identified along a reading line; the diffused light is picked up by a photosensor which generates, in its turn, an analog signal proportional to the intensity of the diffused light.
Said analog signal is processed by a block called digitaliser, whose task is to extract from the analog signal the numeric characteristics that represent it completely, so as to define a scanning of the code being examined.
Such scanning is then elaborated by a bar code decoding programme, which, based on a protocol for decoding the particular type of code, classifies the individual characters one at a time, starting from a synchronism character of start until a synchronism character of stop is reached, so as to provide as a conclusion the set of characters that define the decoded symbol of the bar code being examined.
One of the basic limits of the decoding process described hereinabove lies in that, if the label carrying the code to be decoded shows some alterations (such as, for instance, spots, abrasions or smears) on some points wherethrough the scanning is performed, the decoding process stops as it cannot decode the altered characters of the code, and gives an error indication. In this case, all the information obtained from the processing of said scanning are rejected and a new scanning process, carried out on a different zone of the label, is tried, until a complete string of decoded characters is obtained.
However, such process cannot ensure the decoding of deteriorated or locally altered labels, due to the impossibility of finding an entirely clean scanning.
Local alterations of labels cause reading errors that may be generally classified into three categories:
split-like error: this happens when an individual element (bar or space) is digitalised as if it were at least three smaller elements (a typical error caused by a small width abrasion or spot on the label);
merge-type error: this happens when at least an individual element is ignored during the digitalisation, with ensuing merging of the same with the two adjoining elements into one only element (a typical error caused by a large width abrasion or spot on the label);
misread-type error: this happens when, along a given scanning line, an individual element is digitalised in a distorted manner, altering the classification to a significant extent (a typical error caused by a print smear).
The aforementioned errors may be found also at the same time on a same scanning line of the code being examined, stopping or in any case invalidating the decoding process of the code along such particular scanning, so that the analysis of a subsequent code scanning becomes necessary. In order to produce a feasible and/or quicker decoding scanning, several processes have been proposed in the last years, wherein the code decoding is performed by utilising information taken from several scannings.
For instance, the U.S. Pat. No. 5,262,626 discloses a method for decoding a bar code symbol wherein the information obtained from two different scannings are compared to one another in order to process a complete decoding string. More particularly, the elements of the undecoded characters of a first scanning are iteratively replaced by the elements of the decoded characters of a second scanning having the same position within the code being examined, obtaining in this way an increasingly complete decoding string. If two compared scannings should have, in the same position, two different decoded characters, the known art suggests that the scanning containing the highest number of decoded characters be kept. Besides, if both scannings should have the same decoded characters, the known art suggests that the most recent scanning be kept and a new scanning acquired.
Even though it satisfies the need of accelerating the decoding process, the above described process has several drawbacks.
A first drawback is related to the high computation effort associated to the fact of working with the individual elements of the characters.
A second drawback is related to the low reliability of such decoding process. Actually, the decoded characters are submitted neither to controls nor to exactness checks: this may lead to the processing of a complete but not correctly validable decoding string or, still worse, to a string that, while being validable, does not correspond to the original string.
More particularly, if both of the compared scannings have two different decoded characters in the same position, the decision of considering as “good” the character belonging to the scanning containing the highest number of decoded characters does not provide any warranty as concerns the real validity and/or exactness of said character. In this case it may happen that a character be identified that actually does not exist in the label being examined, attributing therefore to said label a wrong code symbol. Such drawback is partly overcome by U.S. Pat. No. 5,438,188, which teaches a decoding process wherein the classification of the individual characters of a scanning is performed by comparing the elements of the code being examined with all the possible characters defined in the protocol of such particular type of code; an error coefficient is calculated for each comparison and the character with the lowest error coefficient is classified. The decoding of the bar code symbol is carried out thereafter by comparing the information iteratively obtained from two different following scannings, and considering as “good” the decoded character that has, in that particular position, the lowest associated error coefficient.
Such decoding technique, while being an improvement as concerns the reliability of the decoded characters, has still the drawback of requiring high calculation resources for the processing of the individual scannings. Such drawback is due to the necessity of carrying out, for each element of the code, a comparison with all the possible characters of the decoding protocol. Therefore, this process requires a particularly exacting computation effort, which often translates into long decoding times.
In fact, it should be kept in mind that the calculation resources necessary for the decoding and the decoding time are parameters of basic importance for the selection of the decoding programme to be utilised and limit often the application field of said programme.
For instance, the calculation resource parameter is particularly important in case of utilisation of code scanning devices of the manual type, which, for reasons of cost, have limited storing availability and calculation power.
Instead, the time factor is particularly important especially in the cases where—as for instance in shop counters or product storehouses—the identification of the code must take place as quickly as possible, not to create a bottleneck to the flow of goods to be identified.
A further drawback of all the aforementioned decoding processes concerns the impossibility, for the decoding programmes, of completing the reading of the scanning each time an undecodable characters is found in the same.
In s
Bottazzi Marco
Brunelli Elena
Zocca Rinaldo
Datalogic S.p.A.
Frech Karl D.
St.Cyr Daniel
Sughrue Mion Zinn Macpeak & Seas, PLLC
LandOfFree
Process for scanning a bar code does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Process for scanning a bar code, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Process for scanning a bar code will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2600338