SCANNING METHOD FOR A LARGE-SIZE SCANNER SYSTEM USING A STITCHING PROCESS

• US 20140036323A1
• Filed: 04/19/2012
• Published: 02/06/2014
• Est. Priority Date: 04/23/2011
• Status: Active Grant

• Current Assignee: Roth+Weber Gmbh
• Original Assignee: Roth+Weber Gmbh
• Inventors: Kaempflein, Simon, Schoelpen, Burkhard
• Granted Patent: US 8,824,023 B2
• Time in Patent Office: Days
• US Class Current: 358/482
• CPC Class Codes: H04N 1/04 Scanning arrangements , i.e. arrangements for the displacement of active reading or reproducing elements relative to the original or reproducing medium, or vice versaH04N1/387 takes precedence
• CPC Class Codes: H04N 1/1903 Arrangements for enabling electronic abutment of lines or areas independently scanned by different elements of an array or by different arrays
• CPC Class Codes: H04N 1/1933 Staggered element arrays, e.g. arrays with elements arranged in a zigzag
• CPC Class Codes: H04N 1/3876 Recombination of partial images to recreate the original image
• CPC Class Codes: H04N 1/4092 Edge or detail enhancement

Abstract

The invention relates to a scanning method for a large-size scanner system (1) comprising at least two image detection elements (2) that are used for scanning a large-size original (4) to be scanned and are arranged in a cascaded manner so as to have at least one overlapping area (13, 14). Said method, in which the image data in the at least one overlapping area (13, 14) of the image detection elements (2) is combined using a stitching process, said at least one overlapping area (13, 14) being searched for image data within a search region (10), includes the following steps: recognizing (18) the texture within the defined search region (10) of the original (4) to be scanned; assessing the data density in the recognized texture to determine a measure (19) of the texture content; weighting (22) the data density as a function of the measure (19) of the texture content of the recognized texture; detecting (20) congruent image elements within the defined search region (10) of the original (4) to be scanned; for each measured point (16 to 17), determining (24 to 25) a weighted deviation from the weighting (22) derived from the texture and the determined deviation (24 to 25) of each measurement; determining (27, 28) a weighted average of the deviations (29 to 30) from said weighted deviations (24 to 25); and calculating, from said weighted average (27, 28) of the deviation, shift values (29, 30) for correcting the position of the offset image elements such that said image elements are made to be congruent.

Claims

1. A scanning method for a large-format scanner system, having at least two image sensors used for scanning a large-format original and cascaded so as to have at least one overlap region, the image information in the at least one overlap region of the image sensors being combined by a stitching process, the at least one overlap region within a search area being searched for image information, the method comprising the following steps:

S1) recognizing the texture within the defined search region of the original being scanned;

S2) evaluating the information density in the recognized texture to determine a criterion for the texture content;

S3) weighting the information density as a function of the criterion for the texture content of the recognized texture;

S4) detecting a deviation in congruent image information within the defined search region of the original being scanned;

S5) determining a weighted deviation of each measurement point from the weighting derived from the texture and the detected deviation of each measurement, and

S6) determining a weighted average of the deviations from these weighted deviations; and

S7 using this weighted average to calculate the deviation of shift values in order to correct the position of the offset image elements so that these image elements can be made congruent.

2. The scanning method according to claim 1, wherein the size of the defined search area in step S1 can be modified, in particular, enlarged when starting the scan of original being scanned.

3. The scanning method according to claim 1, wherein the detection of congruent image elements in step S4 is done by measuring the determined deviations in the x and y axes.

4. The scanning method according to claim 1, wherein the detection of congruent image information is done in step S4, in addition to texture recognition, at predetermined intervals within the defined search area.

5. The scanning method according to claim 1, wherein the following steps are performed prior to step S1 when starting to scan the original being scanned:

defining a larger search region to detect the leading end of the original being scanned; and

determining a texture value within the search region and weighting the recognized texture.

6. The scanning method according to claim 1, wherein the following additional steps are performed:

determining fine regular information structures of supposedly congruent image information that cannot be uniquely assigned;

comparing the supposedly congruent image information that has been determined in the search area with static stitching parameters that result from the x and y offset of the image sensors relative to each other;

weighting the supposedly congruent image information based on the match with the static stitching parameters; and

detecting image information with the highest weighting as being congruent image information.

7. The scanning method according to claim 6, wherein existing x or y offsets are compensated by correction algorithms.

8. The scanning method according to claim 1, wherein the weighting of the information density for the deviations and selection of the maximum in step S3 are calculated using static stitching parameters.

9. The scanning method according to claim 1, wherein the determination of a weighted average of the deviations in step S6 is done by filtering or averaging while utilizing each weighting of the weighted deviations.