JP3045777B2 - Method and apparatus for evaluating the performance of a printer or imaging system - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to two-dimensional symbols, and more particularly to a method and apparatus for characterizing the quality of two-dimensional symbols in a pixel image.

Description of the Related Art Chandler, et al., US Pat. No. 4,874,
No. 936 and U.S. Pat. No. 4,896,029 to Chandler et al. Disclose information with multiple polygonal elements and an acquisition target (acqu.
article for encoding polygon information, which is encoded in the form of a two-dimensional symbol (figure) composed of
A method and system are described.

U.S. Pat.No. 4,874,936 and U.S. Pat.
The two-dimensional symbol described in 6,029 is a snake-eye pattern (bull'S
−eye pattern) acquisition target
And an array of hexagonal elements with different shading in which the alphanumeric data is encoded. Such a symbol is one inch (2.
About 80-100 pixels (about 32-40 pixels / c) per 54cm
m) can be printed using a printer having a resolution of m). Each hexagonal element of the symbol at this resolution is described in US Pat. No. 4,896,029.
It approximates a true hexagon as shown in Figures 13, 14 and 15 of the issue. Symbols having elements that more or less approximate a true hexagon can be printed on the printer, even if the resolution of the printer is higher or lower. However, as the degree of approximation to a true hexagon decreases (decreases), the number of errors (errors) generated when decoding information stored (stored) in a symbol increases. Therefore, the resolution of the printer used directly affects the symbol decoding performance of the symbol reader.

Further, symbols may be distorted during the printing process. Also, distortion in the image of the printed symbol (ie, the pixel image of the symbol obtained by reading the printed symbol) is caused by the optical sensor system used to generate the image of the printed symbol. Sometimes it happens. Further, such a distortion may cause a decoding error (error) in decoding a printed symbol.

SUMMARY OF THE INVENTION In a preferred embodiment of the present invention, the present invention is a method and apparatus for evaluating the performance of a device (determining performance characteristics). There, a printed symbol containing a control pattern is used (supplied). The printed symbols are those printed by the printer. First, an imaging system (ie, a system for reading as an image,
Image reading system or image input system)
A pixel image of the printed symbol is generated.
The pixel image includes a control pattern that has been captured (that is, read or input as an image). Next, the position of the captured control pattern in the pixel image is obtained. A quality measure is calculated which is a function of the captured control pattern and the representation of the prescribed control pattern. According to the quality measure, the performance of either the printer or the device of the imaging system is evaluated.

In another embodiment of the method and apparatus for evaluating the performance of an apparatus according to the present invention, symbols printed on a printer are used. First, a pixel image of a printed symbol is generated using an imaging system. Next, the position of the captured symbol in the pixel image is obtained. A histogram of at least a part of the captured symbol is generated. A quality measure is calculated according to the histogram. According to the quality measure, the performance of either the printer or the device of the imaging system is evaluated.

In yet another embodiment of the method and apparatus for evaluating the performance of an apparatus according to the present invention, symbols printed on a printer are used. First, using the imaging system,
A pixel image of the printed symbol is generated. Next, the position of the captured symbol in the pixel image is obtained. At least a portion of the imaged symbol is converted to the frequency domain. A quality measure is calculated according to the transformed part. According to the quality measure, the performance of either the printer or the device of the imaging system is evaluated.

One object of the present invention is to provide a method and apparatus for evaluating the performance of a printer used to print two-dimensional symbols.

Another object of the present invention is to provide a method and apparatus for evaluating the performance of an imaging system used to generate a printed two-dimensional symbol image (pixel image).

BRIEF DESCRIPTION OF THE DRAWINGS Other objects, features and advantages of the present invention will become apparent from the following detailed description of the preferred embodiments, the appended claims and the accompanying drawings.

FIG. 1 shows U.S. Pat. No. 4,874,936 and U.S. Pat.
1 shows a two-dimensional symbol described in Japanese Patent Application No. 96,029.

 FIG. 2 shows an ideal hexagonal coordinate system.

FIG. 3 shows an array of hexagons arranged so that each point coincides with each point on an ideal hexagonal coordinate system.

FIG. 4 shows a process flow diagram of a system for evaluating the quality of the symbol of FIG. 1 in a pixel image, according to a preferred embodiment of the present invention.

 FIG. 5 shows a capture target for the symbols of FIG.

FIG. 6 shows a histogram of an 8-bit grayscale pixel image of the symbol of FIG.

FIG. 7 shows a portion composed of only hexagons in the symbol of FIG.

FIG. 8 shows an example of the structure of FIG. 7 using the fast Fourier transform algorithm.
Is a result of converting a part of the frequency domain into the frequency domain.

DETAILED DESCRIPTION OF THE INVENTION The present invention is directed to U.S. Pat. No. 4,874,936 (Chandler (C
(Chandler et al.) and U.S. Pat. No. 4,896,029 (Chandler et al.) to a method and apparatus for evaluating the quality of two-dimensional symbols in pixel images. The entire description of the above-mentioned U.S. Pat. Nos. 4,874,936 and 4,896,029 is incorporated herein by reference. Using such an evaluation method, the quality of the printer used to print the symbol, the quality of the optical sensor system used to generate the image of the printed symbol (pixel image), or both. Quality can be determined or evaluated.

Referring now to FIG. 1, which illustrates the above-mentioned U.S. Pat.
A two-dimensional symbol 100 is shown as described in US Pat. No. 4,874,936 and US Pat. No. 4,896,029. The two-dimensional symbol 100 is composed of an array of capture targets 102 and hexagons 104. In one embodiment, the information comprises a plurality of hexagons 104 with relative shading, such as white or black.
In the form of a symbol 100 expressed in the form The capture target 102 is a kind of control pattern that is preferably located at the center of the symbol 100, and is formed by a snake pattern consisting of three black rings 106 and three white rings 108. The white circle in the center is the above three white rings
One of 108. The control pattern may be any known pattern as long as it does not differ (is not changed) for each symbol. Each capture target 102 is ideally the same for each symbol 100. Therefore, the capture target 102 is a control pattern.

There are several features or characteristics of the image (pixel image) of the two-dimensional symbol 100 used to determine the overall quality of the symbol 100. For example, each array of hexagons 104 is different (varies) for each symbol
However, the capture target 102 is the same for each different symbol 100. The ideal symbol 100 has a defined or defined capture target 102. Therefore, one of the viewpoints regarding the quality of the imaged symbol 100 is as follows.
One is how well the captured target 102 of the captured symbol 100 matches the specified capture target 102 of the ideal symbol 100 (match, match). The higher the degree of coincidence (consistency), the higher the quality of the imaged symbol 100.

Another aspect regarding the quality of the imaged symbol 100 relates to the color or coloring of the pixel image. For example, in an ideal 8-bit grayscale image of symbol 100, black hexagon 104 or black ring 106
Each pixel corresponding to pixel intensity level 0 (zero)
And each pixel corresponding to a white hexagon 104 or white ring 106 has a pixel intensity level 255. In the actual 8-bit grayscale image of the symbol 100, the intensity levels of the "black" pixels may be at different levels other than 0 (zero). Similarly, the intensity levels of the "white" pixels may have different values in a range of values other than 255.

As a result, the quality of the actual captured symbol 100 is determined by evaluating the difference in intensity level (ie, contrast) or separation (distance) between the black and white pixels. The higher the intensity level contrast or separation between the black and white pixels, the higher the quality of the symbol 100. The quality of the symbol 100 is also determined by evaluating the uniformity (uniformity) or variation (change) of the intensity level of each pixel in one of the white or black level ranges. Each pixel within the white area is highly uniform,
The smaller the variation, the higher the quality of the symbol 100. The same applies to black pixels.

Yet another aspect of the quality of the imaged symbol 100 relates to the location of the hexagon 104. The center of each hexagon 104 in the ideal symbol 100 coincides with a certain point in the ideal hexagonal coordinate system. Referring now to FIG. 2, an ideal hexagonal coordinate system is shown. Referring now to FIG. 3, there is shown an array of hexagons 104 arranged to coincide with each point of the ideal hexagonal coordinate system. In an ideal hexagonal coordinate system, each interior point
302 (the point corresponding to the center of the hexagon 104) is the six neighboring points 304 (the point corresponding to the center of each of the six neighboring hexagons 104)
Having. The six neighboring points 304 define an ideal regular hexagon 306 having six equal length sides 308 and six equal angles (ie, 120 degrees) interior angles 310. Therefore, the quality of the imaged symbol 100 is the hexagon 1 in the hexagonal array.
It is determined by evaluating the position of each center of 04.

Each of these different perspectives on the quality of the imaged symbol 100 is a different quality measure (measur) designed to assess the quality of the symbol in different quality perspectives.
e) can be quantified. The overall quality of the imaged symbol 100 can be evaluated based on these different quality measures. For example, a test is performed for each different aspect of symbol quality by comparing the quality measure for each different aspect to a particular threshold of the corresponding quality measure. Then, the overall quality of the imaged symbol 100 is determined based on the results of the different tests. The imaged symbol 100 can be said to have sufficient quality if all or a predetermined number or set of tests pass (pass). Stated another way, a single overall (overall) symbol quality measure is also a function of the different quality measures for each different symbol quality perspective. This global symbol quality measure is compared to a global symbol quality threshold to determine whether the captured symbol 100 is of sufficient quality.

Referring now to FIG. 4, there is shown a process flow diagram of a system 400 for evaluating the quality of a symbol 100 in a pixel image according to a preferred embodiment of the present invention. First, means 402 in system 400 generates a two-dimensional grayscale pixel image of symbol 100. Means 402 is available from Pulnix America, Inc. of Sunnyvale, California 94086, USA.
x America, Inc.), a CCD (i.e., charge coupled device) linear (line) array camera, such as a Pulnix 7-CN black and white CCD camera.

The means 404 then determines the position of the capture target 102 of the symbol 100 in the pixel image. In a preferred embodiment, the means 404 comprises convoluting the pixel image with one or more templates corresponding to a defined capture target.
ion) to determine the position of the capture target 102. In a preferred embodiment, the means 404 comprises the aforementioned U.S. Pat.
As described in US Pat. No. 4,936 and U.S. Pat.
Find the position of 2. In another embodiment, the 1993 3
Joe Zheng filed on January 1
And Jiansu Lai in U.S. Patent Application "Method and Apparatus for Locating Two-Dimensional Symbols Using Double Templates"
Locating a Two−Dimensional Symbol Using a Double
Template) ”(U.S. Postal Service Express
The location of the capture target 102 is determined using a dual template, as described in the Mail Service "Application No. HB200235244".

Also, after determining the location of the capture target 102, the means 404 preferably decodes the symbol 100 to verify or verify that it is a valid symbol. If the symbol 100 is valid, the means 406 evaluates the quality of the captured target. Means 406 will be described in detail later in this specification in connection with FIG.

Means 408 then generates a histogram of pixel intensity levels corresponding to the captured symbols. Then, means 410 calculates some quality measure based on the histogram of the imaged symbols. Means 410
Will be described in detail later in this specification with reference to FIG.

Means 412 then applies a Fourier transform to the pixel intensity levels representing the captured symbols to convert the intensity level values to the frequency domain. Means 414 then calculates some quality measure for the hexagonal position of the imaged symbol. Means 414 will be described in detail later in this specification in connection with FIGS.

Means 416 then generates an output of the results of the quality measure of the captured capture targets and symbols. Preferably, the output is in the form of a screen display and / or a hard copy printout, depending on the user's selection.

Assessing the Quality of Capture Targets Referring now to FIG. 5, a capture target 102 for a symbol 100 is shown. The capture target 102 comprises a serpentine pattern consisting of three black rings 106 and three white rings 108. The white circle at its center is one of the three white rings 108. According to a preferred embodiment of the present invention, a quality measure of the captured capture target can be determined by analyzing a grayscale image representing the capture target. This grayscale image is thresholded to generate a binary image representing the capture target. Then, the binary image is run-length (run-lengt).
h) encoded.

In a preferred embodiment, the grayscale image of the captured target is thresholded using a global threshold GT. If you are an expert in this field,
You can see that GT can be selected in several ways. In a preferred embodiment, the global threshold GT is determined by the entropy criterion or standard (crit) given by equation (1)
erion) Φ (s).

Here, p _s represents the probability of occurrence of gray level s appearing in the gray scale image is determined from Pisutoguramu of the image. E _s is the entropy at the gray level s, and the gray level takes a value from 0 to M (the maximum gray level in the image). The entropy E _{s at the} gray level s is given by the following equation (2).

Here, p _i is the appearance probability of gray level i. The entropy criterion Φ (s) is calculated for each gray level value s (s = 0 to
M). Next, the global threshold GT
Is set equal to the value of s corresponding to the maximum value of the entropy criterion Φ (s).

The pixel image is thresholded by comparing each pixel intensity level value with a global threshold GT. When the pixel intensity level value is larger than the global threshold value GT, a value “1” representing a white pixel is designated as the thresholded pixel. In other cases, a value “0” representing a black pixel is designated as the thresholded pixel.

In run-length encoding, a sequence of pixels in a binary pixel image is represented by its run-length encoded value. For example, FIG. 5 is a binary image representing a captured target 102, the white pixels of which are represented by "1",
Assume that the black pixel is represented by "0". In a preferred embodiment, the binary image of FIG. 5 is generated by thresholding a grayscale image. A row of pixels 502 through the center of the capture target 102 has the next pixel sequence in the left-to-right direction starting with the first black pixel.

Here, each black ring 106 and each white ring 108 has a width of five pixels in the pixel image. According to an embodiment of the present invention, pixel row 502 is represented by a run-length encoding of the pixel sequence as follows:

Where the first entry (10) is the number of transitions from black to white and from white to black along row 502;
Each successive number (5) is the number of consecutive white or black pixels in pixel row 502 starting from the first black pixel in the left-to-right direction.

To give another example, pixel row 504 contains the next pixel
It may have a sequence.

According to the present invention, pixel row 504 is represented as a run-length encoded sequence as follows.

Here, there are six transitions from black to white or from white to black, and the sequence of consecutive black and white pixels is as described above.

The means 406 converts the binary image of the captured target 102 into a grayscale "converted image" using the run-length encoding method. Each row of the transformed image is composed of a value resulting from run-length encoding of the corresponding row of the binary image.
The transformed image has the same number of rows as the binary image, but generally
It has fewer columns than the value image. Furthermore, the number of pixels in each row of the transformed image generally varies from row to row (varies). In the above example, the transformed row 502 has 12 pixels, but the transformed row 504 has only 8 pixels.

Means 406 determines the quality of the captured image (the position determined by means 404) by run-length encoding the transformed image and the specified capture target (generated off-line) by a run-length encoded representation. Generate a scale. In a preferred embodiment of the present invention, the quality measure is a correlation coefficient ρ (i, i) between the two run-length coded images.
j). This correlation coefficient ρ (i, j) is determined using the following equation (3).

Where C _ft (i, j) is the cross-correlation or cross-variance (cross-correlation) between the run-length coded imaged capture target and the run-length coded defined capture target. variance) and σ _f (i, j) are the autocorrelations (auto) of the run-length coded imaged captured target.
-Correlation) or auto-variance
, And σ _t represents the autocorrelation or self variance of the run-length coded defined acquisition target. C _ft (i, j), σ
_f (i, j) and σ _t are determined according to the following equations (4) to (6).

Where f (x + i−M / 2, y + j−n / 2) is the row x of the run-length encoded captured target pixel image
+ I−M / 2 and the pixel intensity level at column y + j−n / 2. (I, j) is the run-length coded average pixel intensity level value of the captured capture target. t
(X, y) is the pixel intensity level value in row x and column y of the run-length coded defined capture target image.
Also, is the average pixel intensity level value of the specified capture target run-length encoded. Furthermore, (i, j)
And are determined according to the following equations (7) and (8).

Here, M represents the number of rows of the run-length encoded target capture target image, and N represents the number of columns thereof. (I, j) is
It represents the position of the captured target in the run-length encoded transformed image.

The correlation coefficient ρ (i, j) takes a value between −1 and 1. Here, the value 1 corresponds to a perfect match between the defined capture target and the captured capture target. Therefore, the correlation coefficient ρ
The greater (i, j), the higher the match between the defined capture target and the captured capture target, and the higher the image quality of the captured capture target in this image quality perspective. Therefore, the correlation coefficient ρ (i, j) (the cross-correlation C _ft (i,
j) but not) to obtain a relative measure of the quality of the imaged capture target relative to the defined capture target.

The use of run-length coded images reduces the dimensions of the images used in the calculation, thereby reducing the computational complexity in calculating the correlation coefficient ρ (i, j).

In addition to the correlation coefficient ρ (i, j), the horizontal and vertical dimensions of the captured target are determined. The horizontal dimension of the imaged capture target is the distance across (passing) the capture target along a row of pixels passing through the center of the capture target.
The vertical dimension of the captured capture target is the distance across the capture target along a column of pixels passing through the center of the capture target. In an ideal symbol, the horizontal and vertical dimensions are equal to each other and to the dimensions of a defined capture target. Any differences between these dimensions indicate errors in the printing process, for example, the paper feed mechanism of the printer.
It indicates that there is.

Experts in the art will appreciate that the system 400 of the present invention
It will be appreciated that the design can also be designed to evaluate the quality of control patterns other than the capture target 102 of the symbol 100.

Histogram-Based Quality Measure As described earlier herein in connection with FIG. 4, means 408 in system 400 generates a histogram of pixel intensity level values of the grayscale pixel image of symbol 100. Means 410 then calculates a quality measure based on the histogram that evaluates a quality characteristic of the pixel image in terms of that quality.

Next, referring to FIG. 6, the symbol 100
Is shown for the 8-bit grayscale pixel image. As shown in FIG. 6, the image pixels are clustered into a set of two intensity level regions. One region 602 corresponds to a black pixel, and another region 604 corresponds to a white pixel. The black reflectance value (BR) is the pixel intensity level corresponding to the peak (maximum value) of the black pixel area 602 (ie, the black pixel intensity level that appears most frequently in the pixel image). Similarly, the white reflectivity value (WR) is the pixel intensity level corresponding to the peak of the white pixel area 604 (ie, the most frequently occurring white pixel intensity level in the pixel image). The global threshold (GT) generated using equation (1) above is illustrated in FIG. This global threshold can be used to discriminate between white and black pixels in the histogram of FIG.

In a preferred embodiment, the means 410 calculates the symbol contrast value SC from the values of the white reflectivity WR and the black reflectivity BR according to the following formula:

The symbol contrast value SC indicates the contrast between the reflectivity of the black and white pixels. The larger the symbol contrast value SC, the higher the image quality of the symbol 100 with respect to the symbol quality in terms of pixel contrast.

Further, the means 410 calculates the value ERU of the uniformity (uniformity) of the element reflectivity for both black and white pixels according to the following equations (10) and (11).

Here, H (BR) and H (WR) are histogram values of the black reflectivity value BR and the white reflectivity value WR, respectively. The denominator in equation (10) is the sum of the histogram values H (i) at the intensity level value i of the black pixel (ie, a black pixel having an intensity level value less than or equal to the global threshold GT). Similarly, the denominator of equation (11) is the sum of the histogram values H (i) at the intensity level value i of the white pixel (ie, a white pixel having an intensity level value greater than the global threshold GT). Value ERU _b and ERU elements reflectance uniformity _w indicates the uniformity of the black and white pixels, respectively. The higher the value of the element reflectivity uniformity, the higher the image quality of the symbol 100 in terms of pixel uniformity. The ideal pixel image of symbol 100 has an element reflectivity uniformity value of one.

Further, the means 410 calculates the element reflectivity variance (ERV) for the black and white pixels according to the following equations (12) and (13).

Here, p _i is the appearance probability of the pixel intensity level i, and is obtained according to the following equation (14).

_b and _w are the average values of the black and white pixels, respectively, and are calculated according to the following equations (15) and (16).

Element reflectance variance ERV _b and ERV _w each represent the variance of each pixel in the range of black and white. The smaller the variance of the element reflectivity, the higher the image quality of the symbol 100 in terms of pixel variance. The ideal pixel image of symbol 100 has an element reflectivity variance of 0 (zero). The element reflectivity variance can indicate how the tone changes during printing of the symbol 100 by the printer.

Also, the means 410 calculates the minimum reflectivity difference MRD according to the following equation (17).

Here, as shown in FIG. 6, (WR-ERV _w ) is defined as the intensity level of dark white pixels, and (BR + ERV _b ) is defined as the intensity level of light black pixels. BR
And WR are reflectance values of black and white, respectively, a black and a value of the uniformity of element reflectance of white according to the ERV _b and ERV _w is each of the aforementioned formula (12) and (13).

The minimum reflectivity difference MRD indicates the degree of separation between black and white pixels. The higher the value of the MRD, the higher the image quality of the symbol 100 in terms of pixel separation. In an ideal pixel image of the symbol 100, ERV _b and ERV _w are both 0 (zero), the minimum reflectance difference MRD is 1.0.

Hexagonal Position Quality Measure As discussed earlier in this specification in connection with FIGS. 2 and 3, the quality perspective of the imaged symbol 100 is determined by the location or position of the center of the hexagon 104 in the hexagonal array. Also relevant. The system 400 uses the Fourier analysis to
Determine a quality measure of the placement or location of As described earlier herein in connection with FIG.
The inner means 412 converts portions of the grayscale pixel image of the symbol 100 that do not include the capture target 102 (ie, portions of the pixel image of the symbol 100 that consist solely of the hexagons 104). The means 104 is a fast Fourier transform (FFT)
The algorithm converts only the hexagonal portion of the pixel image into the frequency domain. Means 414 uses the transform result of the FFT to calculate a quality measure of the hexagonal location indicating the location of the center of hexagon 104 in the hexagonal array.

Referring now to FIG. 7, there is shown a portion of the pixel image of symbol 100 that is comprised solely of hexagons 104. 7, line 1 _1, 1 ₂ and 1 ₃ correspond to the three axes of the hexagonal array, the angle alpha _1, alpha ₂ and alpha ₃ corresponds to the angle between the respective axes, the distance D _1, D ₂ And D ₃ is hexagon 10
4 corresponds to the distance between the centers in the direction perpendicular to each axis. For example, the angle alpha ₁ is the angle between lines 1 ₁ and 1 _2, the distance D ₁ is the distance in the direction perpendicular to the line 1 ₁ between the centers of the hexagon 104. In an ideal hexagonal perfect pixel image, the angles α ₁ , α ₂ and α ₃ are all 60 degrees and the distances D ₁ , D ₂ and D ₃ are all equal.

Referring now to FIG. 8, there is shown a graphical representation of the result of transforming an ideal hexagonal complete pixel image into the frequency domain using the FFT algorithm.
In that frequency domain, a complete hexagonal pixel image positioned to coincide with each point of the hexagonal coordinate system has three pairs of equivalents of the frequency of change inherent in the hexagonal grid of symbols 100. Represented by points p ₁ , p ₂ and p ₃ . Angle alpha ₁ between the lines connecting p ₁ point 3 pairs, p ₂ and p ₃ of each pair, _{respectively,} alpha ₂ and alpha _3, the angle alpha _1, shown in Figure _7, the same as the alpha ₂ and alpha ₃ Angle. For example,
Angle alpha ₁ is the angle between the line connecting the lines and points p ₂ connecting the points p _1. Errors (deviations, deviations) in the values of the angles α ₁ , α ₂ and α ₃ from 60 degrees indicate imperfections in the arrangement of the hexagons 104 in the pixel image of the printed symbol 100. The error is called the element angle (EA) error and is calculated by the means 414.

Further, the means 414 calculates the error of the element arrangement EP in the frequency domain based on the distances d ₁ , d ₂ and d ₃ . distance
d ₁ and d ₂ and d ₃ is the distance of each from the origin in the frequency domain to each point _{p 1, p 2, p 3} . The distances d ₁ , d ₂ and d ₃ in the frequency domain can be converted to the spatial domain by dividing by the number of samples N used in the FET algorithm. Next, the converted distance is compared with the distances D ₁ , D ₂ and D ₃ in FIG. 7 to determine the error of the element arrangement EP with respect to each axis. The smaller the element placement error, the hexagonal geometry (geometry)
The image quality of the symbol 100 from the viewpoint of is higher.

Those skilled in the art will appreciate that the element angle EA and element arrangement EP measures are global quality measures that determine the overall distortion of the arrangement of the hexagons 104 of the symbol 100. The misplacement of a small number of hexagons 104 relative to the total number of hexagons 104 analyzed has little effect on these quality measures.

Evaluating Printer Performance According to the present invention, it is possible to determine whether a symbol printed by the printer satisfies a predetermined minimum print quality by evaluating the performance of the printer. Thus, the present invention can be used to analyze the printing performance of any commercially available (or custom) label printer. Examples of such label printers are the UBI printer 201 of the UBI printer AB in MoIndal 22, S-431, PO Box 123, Sweden.
In such applications, it is preferable to reduce and / or take into account (compensate for) any effect of the optical sensor system on the quality of the image obtained from the printed symbols. The optical sensor system in the preferred embodiment is a CCD-type linear array camera, such as a Pulnix 7-CN black-and-white CCD camera from Prunix America of Sunnyvale, 94086, California, USA.

For example, an optical sensor introduces (gives) random noise to the pixel image of the printed symbol. This image noise may be caused by noise in the CCD array and / or CCD
It can result from noise in the video transmission line carrying the output of the array. In a preferred embodiment, the system
The 400 means 402 reduces this noise by generating the plurality of pixel images from a single printed label and averaging them. If the noise is random and the autocorrelation is 0 (zero), the averaging process improves the signal-to-noise ratio (S / N ratio) of the image. In a preferred embodiment, the means 402 averages four to eight pixel images to form an average image of the symbol 100.

The effect that an optical sensor has on the image obtained from a printed symbol is known as optical distortion, for example, optical distortion due to optical lens errors and / or imperfect dimensions of the CCD array. In a preferred embodiment of the present invention, the means 402 corrects the optical distortion generated by the optical sensor by transforming the image according to a predetermined transformation method. By analyzing the image for a known pattern and mathematically modeling the optical distortion off-line, the transformation method used to remove the effects of optical distortion can be formed. For example, rectangular grid points separated by a known distance can be used as the known pattern. The transformation can be formed by analyzing the distance between points in an image of a known pattern generated by an optical sensor.

In another embodiment, the means 406 uses pre-distorted defined capture targets to determine the quality of the captured capture targets. That is, the means 406 does not modify the image obtained from the capture target of the printed symbol in response to the known distortion caused by the optical sensor, but instead performs a pre-distorted off-line pre-distortion according to the known distortion. Use capture targets. By eliminating known distortions from the captured capture targets without performing the conversion in real time,
6. The processing time can be reduced. As described earlier in this specification in connection with FIG. 5, the captured capture target and the pre-distorted defined capture target are run-length encoded before calculating the correlation coefficient.

Those skilled in the art will appreciate that the present invention can be used to evaluate the performance of an imaging system used to generate an image from a printed symbol. In applying the present invention, it is preferable to reduce and / or take into account (compensate for) any effect of the printer used to print the symbol on the quality of the image obtained from the printed symbol. In evaluating the performance of either device (ie, a printer or an imaging system), the present invention is used to determine the quality of the image obtained from the printed symbols. The user may then determine whether any quality errors (errors indicated by one or more quality measures) are due to the printer used to print the symbol or to generate the image from the printed symbol. It is determined whether it is caused by the imaging system used.

The present invention is directed to a digital signal processing (DSP) chip, such as C-language software running on a Model C-30 DSP chip manufactured by Texas Instruments of Houston, Texas, USA. It is preferable to carry out.

Those skilled in the art will appreciate that features of the present invention can be used to evaluate the quality of symbols other than those described in U.S. Patent Nos. 4,874,936 and 4,896,029. For example, another preferred embodiment of the present invention can be used to evaluate the quality of images obtained from two-dimensional symbols having rectangular or triangular data encoding elements. In addition, features of the present invention can be used to evaluate the quality of one-dimensional symbols, such as barcode symbols.

Those skilled in the art will also understand that quality measures other than those specifically described in this specification are within the scope of the invention. For example, histogram-based quality measures and hexagonal arrangement quality measures other than the quality measures described herein are also within the scope of the invention.
Further, quality measures other than the quality measure based on the histogram and the hexagonal arrangement are also included in the scope of the present invention.

Further, those skilled in the art will understand the details, elements, and elements of the components described and illustrated to explain the features of the invention without departing from the principles and scope of the invention as set forth in the appended claims. Obviously, various modifications of the configuration are possible.

────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Lai, Chien Su, Massachusetts, United States 02173 Lexington Grapevine Avenue 11 (56) References JP-A-6-323814 (JP, A) JP-A-6-233013 (JP) JP-A-6-30161 (JP, A) JP-A-7-193709 (JP, A) JP-A-8-69534 (JP, A) French Patent Application Publication 2704338 (FR, A1) (58) ) Field surveyed (Int.Cl. ⁷ , DB name) H04N 1/00 B41J 29/00 B41J 29/46 G03G 15/00 G03G 21/00

Claims (30)

(57) [Claims] 1. A step of: (a) supplying a symbol printed with a printer including a control pattern; and (b) generating a pixel image including a control pattern of the printed symbol by using an imaging system. (C) determining a position of the captured control pattern in the pixel image; and (d) calculating a quality measure that is a function of the captured control pattern and an expression representing a prescribed control pattern. (E) evaluating the performance of one of the printer and the imaging system according to the quality measure. 2. The method of claim 1, wherein said quality measure comprises a correlation coefficient between said captured control pattern and said representation. 3. A step (d) comprising: (1) run-length encoding the imaging control pattern to generate a run-length encoded imaging control pattern; and (2) the run-length encoded imaging control pattern. Calculating the quality measure as a function of a control pattern and a run-length coded representation of the defined control pattern. 4. The step (1) of the step (d) includes: (i) generating a binary imaging control pattern by performing threshold processing on the imaging control pattern; and (ii) generating a binary imaging control pattern. 4. The method of claim 3, further comprising the step of: run-length-encoding to generate the run-length-encoded imaging control pattern. 5. The method of claim 1, wherein step (d) includes calculating the quality measure as a function of the imaging control pattern and a pre-distorted representation of the defined control pattern. . 6. The step (d) includes: (1) generating an expression representing the specified control pattern; (2) distorting the expression; and (3) run-length converting the distorted expression. Encoding; (4) when the imaging control pattern is not a binary image, generating a binary imaging control pattern by performing a threshold process on the imaging control pattern; and (5) generating the binary imaging control pattern. (6) run-length encoding the imaging control pattern; and (6) comprising a correlation coefficient between the run-length encoded binary imaging control pattern and the run-length encoded distorted representation. The method of claim 1, comprising calculating a quality measure. 7. The method of claim 1, wherein the quality measure is responsive to a distance across the imaging control pattern along a row of pixels passing through the center of the imaging control pattern. 8. The method of claim 1, wherein the quality measure is responsive to distance across the imaging control pattern along a row of pixels passing through the center of the imaging control pattern. 9. A means for generating a pixel image including a control pattern of a symbol printed by a printer using an imaging system; and (b) determining a position of the captured control pattern in the pixel image. Means for calculating; (c) means for calculating a quality measure which is a function of the imaged control pattern and an expression representing a prescribed control pattern; and (d) according to the quality measure, any of the printer and the imaging system. A device for evaluating the performance of the device, comprising: means for evaluating the performance of one device; 10. The quality measure includes a correlation coefficient between the imaging control pattern and the representation.
An apparatus according to claim 1. 11. The means (c) comprises: (1) means for run-length encoding the imaging control pattern to generate a run-length encoded imaging control pattern; and (2) the run-length encoded control pattern. The apparatus of claim 9, comprising: means for calculating the quality measure that is a function of an imaging control pattern and a run-length coded representation of the defined control pattern. 12. The means (1) of the means (c) includes: (i) means for generating a binary imaging control pattern by performing threshold processing on the imaging control pattern; and (ii) the binary imaging control. 12. The apparatus according to claim 11, comprising: means for run-length encoding a control pattern to generate the run-length encoded imaging control pattern. 13. The method of claim 9, wherein said means (c) comprises means for calculating said quality measure which is a function of said imaging control pattern and a pre-distorted representation of said prescribed control pattern. An apparatus according to claim 1. 14. The means (c) includes: (1) means for generating an expression representing the prescribed control pattern; (2) means for distorting the expression; and (3) running the distorted expression. (4) when the imaging control pattern is not a binary image, means for performing threshold processing on the imaging control pattern to generate a binary imaging control pattern; and (5) the binary Means for run-length encoding the imaging control pattern of (a), (6) including a correlation coefficient between the run-length encoded binary imaging control pattern and the run-length encoded distorted representation. The apparatus according to claim 9, comprising means for calculating the quality measure. 15. The apparatus of claim 9, wherein the quality measure is responsive to a distance across the imaging control pattern along a row of pixels passing through the center of the imaging control pattern. 16. The apparatus of claim 9, wherein said quality measure is responsive to a distance across said imaging control pattern along a row of pixels passing through the center of said imaging control pattern. 17. A method comprising: (a) providing a symbol printed by a printer; (b) generating a pixel image of the printed symbol using an imaging system; and (c) imaging in the image. (D) generating a histogram of at least a portion of the imaged symbol; (e) calculating a quality measure according to the histogram; and (f) calculating a quality measure according to the quality measure. Estimating the performance of one of the printer and the imaging system. 18. The method of claim 17, wherein said quality measure comprises symbol contrast. 19. The method of claim 17, wherein said quality measure includes reflectivity uniformity of the element. 20. The method according to claim 17, wherein said quality measure comprises an element's reflectivity variance. 21. The method of claim 17, wherein said quality measure comprises a minimum reflectivity difference. 22. (a) means for generating a pixel image of a symbol printed by a printer using an imaging system; (b) means for determining a position of a captured symbol in the image; Means for generating a histogram of at least a portion of the generated symbol; (d) means for calculating a quality measure according to the histogram; and (e) measuring the performance of one of the printer and one of the imaging systems according to the quality measure. A device for evaluating the performance of the device, comprising: means for evaluating; 23. The apparatus of claim 22, wherein said quality measure comprises symbol contrast. 24. The apparatus according to claim 22, wherein said quality measure includes the reflectivity uniformity of the element. 25. The apparatus of claim 22, wherein said quality measure comprises an element's reflectivity variance. 26. The apparatus of claim 22, wherein said quality measure comprises a minimum reflectivity difference. 27. (a) providing a symbol printed by a printer; (b) generating a pixel image of the printed symbol using an imaging system; and (c) imaging in the image. (D) transforming at least a portion of the imaged symbol into the frequency domain; and (e) calculating a quality measure according to the transformed portion, wherein the quality measure of the element in the frequency domain is calculated. Determining an angle; and (f) evaluating the performance of one of the printer and the imaging system according to the quality measure. 28. (a) providing a symbol printed by a printer; (b) generating a pixel image of the printed symbol using an imaging system; and (c) imaging in the image. (D) transforming at least a portion of the imaged symbol into the frequency domain; and (e) calculating a quality measure according to the transformed portion, wherein the quality measure of the element in the frequency domain is calculated. Determining a position; and (f) evaluating a performance of one of the printer and the imaging system according to the quality measure. 29. (a) means for generating a pixel image of a symbol printed by a printer using an image pickup system; (b) means for obtaining the position of the picked-up symbol in the image; Means for transforming at least a portion of the converted symbol into the frequency domain; (d) means for determining an angle of the element in the frequency domain, and means for calculating a quality measure according to the transformed portion; Means for evaluating the performance of one of the printer and the imaging system according to a scale. 30. A means for generating a pixel image of a symbol printed by a printer using an imaging system; (b) means for determining the position of a captured symbol in said image; and (c) said imaging. Means for transforming at least a part of the converted symbol into the frequency domain; (d) means for determining the position of the element in the frequency domain, and means for calculating a quality measure according to the transformed part; Means for evaluating the performance of one of the printer and the imaging system according to a scale.