Method of auto-cropping images for scanners

• US 5,880,858 A
• Filed: 12/31/1997
• Issued: 03/09/1999
• Est. Priority Date: 12/31/1997
• Status: Expired due to Term

• Current Assignee: Intellectual Ventures Assets 161 LLC (Intellectual Ventures LLC)
• Original Assignee: Mustek Systems Incorporated
• Inventors: Jin, Yuan-Chang
• Primary Examiner(s): BOEHLER, ANNE MARIE M
• Assistant Examiner(s): Grant, II, Jerome
• Application Number:
  US08/001,979
• Time in Patent Office: 433 Days
• Field of Search: 358/465, 358/466, 358/487, 358/474, 358/453, 358/462, 358/467, 348/64, 382/169, 382/172, 382/175, 382/176
• US Class Current: 358/487
• CPC Class Codes: H04N 1/00681 Detecting the presence, position or size of a sheet or correcting its position before scanning H04N1/047 takes precedence; mode signalling H04N1/333; skew detection or correction in image signals H04N1/3878
• CPC Class Codes: H04N 1/00702 Position
• CPC Class Codes: H04N 1/00708 Size or dimensions
• CPC Class Codes: H04N 1/00748 Detecting edges, e.g. of a stationary sheet
• CPC Class Codes: H04N 1/00753 of a sheet relative to a particular backgroud
• CPC Class Codes: H04N 1/00758 using a prescan
• CPC Class Codes: H04N 1/00769 Comparing, e.g. with threshold
• CPC Class Codes: H04N 1/38 Circuits or arrangements for blanking or otherwise eliminating unwanted parts of pictures H04N1/387 takes precedence

Abstract

A method of auto-cropping images appropriate for scanners is disclosed, that features use of the image-division method for carrying out auto-cropping. According to the method, the images of the objects after pre-scanning can be auto-cropped, and therefore do not require further manual operation. The interference due to the background color of the cover and other redundant images can also be reduced. Furthermore, the present invention is also appropriate for scanning positive and negative films, and films are disposed on frames can be scanned properly.

Claims

1. A method of auto-cropping images for scanners comprising the steps of:

a. providing a prescanned image wherein the pixels of said image are processed by an image-division method to obtain at least a low threshold and a high threshold;

b. comparing the pixels in every horizontal row with said low threshold, wherein the number of the pixels that exceed said low threshold is recorded respectively to obtain a dot-intension number for every row, and the dot-intension number of every row is compared with a limit, whereby cropping the rows with a dot-intension number that exceeds said limit, and dividing the prescanned image into several divided image regions;

c. comparing the pixels in every vertical column of every said divided image region with said low threshold, and the number of the pixels that exceed said low threshold is recorded respectively to obtain a dot-intension number for every column, then the dot-intension number of every column is compared with said limit, and the columns with dot-intension numbers that exceed said limit are cropped, and every said divided image region is further divided into several cropped regions; and

d. iterating steps b and c to further divide the cropped regions horizontally and vertically, and stopping the iterating process when every horizontal and vertical division in the cropped regions can not form any new divided region.

2. The method as claimed in claim 1, wherein said image-division method comprises the steps of:

a. selecting a plurality of different initial values with respect to a plurality of sets one to one, serving as the central characteristic values of said sets;

b. calculating the color values of every pixel to obtain the pixel characteristic values corresponding to every pixel;

c. calculating every said pixel characteristic value with said central characteristic values by subtraction and continuously taking the absolute values thereof, whereby selecting the central characteristic value closest to said pixel characteristic value, and by this way, every pixel characteristic value is assigned to the corresponding set of central characteristic values, which is closest to the pixel characteristic value;

d. averaging the pixel characteristic values in every set respectively to obtain corresponding mean characteristic values; and

e. comparing every said mean characteristic value in every set with the corresponding central characteristic value, wherein if every difference is within a specific range, then the minimal mean characteristic value serves as said low threshold, and the maximal mean characteristic value serves as said high threshold, and if at least one of the differences exceeds said specific range, then every central characteristic value of said sets is replaced by said mean characteristic values of said sets and the steps c, d, and e are iterated.

3. The method as claimed in claim 1, when the scanners have covers, said method further comprises the steps of:

a. prescanning a background image and storing it, before putting any objects in a scanner;

b. detecting whether said cover is closed or not, wherein if not closed, then the steps a˜

d in claim 1 are carried out to auto-crop the images, and if closed, then the process comprises the following steps;

i. calculating the background image and a prescanned image by subtraction; and

ii. carrying out the steps a˜

d in claim 1 to the image after subtraction to complete the auto-cropping process.

4. The method as claimed in claim 3, wherein, after carrying out step ii, the margin of said prescanned image is masked out first and then the step ii is carried out, and after completing the cropping of the unmasked region, the cropping regions are expanded toward the masked regions horizontally or vertically to crop the images in the masked regions serving as recovery regions, and the recovery regions are processed by step ii to crop the masked image, thereby completing the cropping of all images.

5. A method of auto-cropping images for scanners appropriate for scanning films with or without frames comprising the steps of:

(I). confirming the type of the films first, and detecting frames, and if frames are detected, then step (II) is carried out;

(II). preprocessing a prescanned image according to the film type, and carrying out the following steps;

2a. processing the pixels of said prescanned image by the image-division method to obtain at least a low threshold and a high threshold;

2b. comparing the pixels in every horizontal row with said low threshold, and recording respectively the number of the pixels that exceed said low threshold to obtain a dot-intension number for every row, then comparing the dot-intension number of every row with a limit, thereby cropping the rows with dot-intension numbers that exceed said limit, and dividing the prescanned image into several divided image regions;

2c. comparing the pixels in every vertical column of every said divided image region with said low threshold, then respectively recording the number of the pixels that exceed said low threshold to obtain a dot-intension number for every column, then comparing the dot-intension number of every column with said limit, thereby cropping the columns with dot-intension numbers that exceed said limit, and further dividing every said divided image region into several cropped regions;

2d. iterating steps 2b and 2c to further divide the cropped regions horizontally and vertically, and stopping the iterating process when every horizontal and vertical division of the cropped regions can not form any new divided region; and

2e. comparing the pixels in every cropped region with said high threshold, and recording the number of the pixels that exceed the high threshold to serve as a sum number, and if said sum number of every cropped region is greater than the total of pixel dots in a cropped region for a certain proportion, then no film is disposed at the cropped position, whereby every cropped region can be checked sequentially to locate the correct position of said film;

wherein if in step I the frames are not detected, then the steps described steps 2a˜

2d in process II are carried out to crop the films regions properly.

6. The method as claimed in claim 5, wherein the frame detection step I comprises the following sub-steps:

1a. processing all pixels in a preview window to obtain a low threshold and a high threshold;

1b. comparing the pixels in every horizontal row with said low threshold, and respectively recording the number of the pixels that exceed said low threshold to obtain a dot-intension number with respect to every row, then comparing the dot-intension number of every row with a limit, thereby cropping the rows with dot-intension numbers that exceed said limit, and recording the row number of every cropped regions;

1c. transforming the size of the regions for disposing films into pixel dots; and

1d. comparing every one of said horizontal row number recorded with said pixel dot-number of the region for disposing a film, and if the difference between at least two of said horizontal row numbers and said dot-number is within a specific range then frames are detected.

7. The method as claimed in claim 5, wherein the image-division method comprises the following steps:

a. selecting a plurality of initial values with respect to a plurality of sets one to one to serve as the central characteristic values of said sets;

b. calculating the color values of every pixel to obtain the pixel characteristic values corresponding to every pixel;

c. calculating every said pixel characteristic value with said central characteristic values by subtraction and continuously taking the absolute values, thereby selecting the central characteristic value closest to said pixel characteristic value, whereby every pixel characteristic value is put into the corresponding set of the central characteristic value, that is closest to the pixel characteristic value;

d. averaging the pixel characteristic values in every set respectively to obtain corresponding mean characteristic values; and

e. comparing every said mean characteristic value in every set with the corresponding central characteristic value, if every difference is within a specific range, then the minimal mean characteristic value serves as said low threshold, and the maximal mean characteristic value serves as said high threshold, and if at least one of the differences exceeds said specific range, then every central characteristic value of said sets is replaced by said mean characteristic values of said sets, then the steps c, d, and e are iterated.

8. The method as claimed in claim 5, wherein, after confirming the type of films in step I, if said films are negative, then said prescanned image is subjected to a reversal operation first, and the frame detection process is carried out successively;

while if said films are positive, then the frame detection is carried out successively without a reversal operation.

9. The method as claimed in claim 5, wherein if the films are negative, then said prescanned image is subjected to a reversal operation, and the cropping process are carried out successively;

while if the films are positive, then the cropping process are carried out to said prescanned image without a reversal operation.