System and method for patterned artifact removal for bitonal images
Images in bitonal formats often include watermarks, stamps, or other patterns and artifacts. These patterned artifacts may be represented as a series of geometric points, dots, and/or dashes in the general shape of the original pattern. These patterned artifacts make other processes such as optical character recognition (OCR) difficult or impossible when items or pixels of interest are also found within the pattern of such artifact(s). Current patterned artifact removal solutions use methods of erosion to minimize the unwanted patterned artifact. However, such methods also erode the pixels/items of interest which, in turn, cause failures in other processes, such as OCR, that are desired to be carried out on or with the image.
1. A method for removing a patterned artifact from an initial bitonal image, the method being performed by one or more processors and comprising:
identifying the patterned artifact in the initial bitonal image,
wherein the identifying the patterned artifact step includes:
generating a modified image by applying an erosion algorithm based on stroke width of the initial image;
marking noise locations of the modified image; and
defining, in the modified image, an artifact boundary and thereby identify the patterned artifact;
removing the patterned artifact to create a new bitonal image with the patterned artifact substantially removed; and
identifying text in the new bitonal image; and
cleaning the new bitonal image to remove one or more remaining portions of the patterned artifact and thereby create a second new bitonal image.
2. The method of claim 1 , wherein the removing the patterned artifact step includes:
applying one or more filters in the artifact boundary defined from the identifying the patterned artifact step;
dilating the identified patterned artifact; and
removing areas having a size larger than areas marked as noise from the marking step of the identifying the patterned artifact step and thereby creating the new bitonal image with the patterned artifact substantially removed.
3. The method of claim 2 , wherein the step of identifying text in the new bitonal image includes:
identifying text in the new bitonal image;
rank reducing the identified text to create one or more potential blocks of text;
a first removing of one or more remaining portions of the patterned artifact through applying a binary AND operation on the identified text from the identifying text in the new bitonal image step with the one or more potential blocks of text from the rank reducing step to create a first intermediate bitonal image of the new bitonal image;
dilating the first intermediate bitonal image; and
comparing the dilated first intermediate bitonal image with the initial bitonal image within the artifact boundary defined from the identifying the patterned artifact step thereby further removing the one or more remaining portion of the patterned artifact and creating a second intermediate bitonal image of the new bitonal image.
4. The method of claim 3 , wherein the cleaning the new bitonal image step includes:
cleaning the second intermediate bitonal image by applying a binary AND operation on the second intermediate bitonal image with the initial bitonal image within the artifact boundary defined from the identifying the patterned artifact step and thereby create the second new bitonal image.
5. A method for removing a patterned artifact from an initial bitonal image, the method being performed by one or more processors and comprising:
(a) identifying the patterned artifact in the initial bitonal image, wherein the identifying the patterned artifact step includes:
generating a modified image by applying an erosion algorithm based on stroke width of the initial image; and
defining, in the modified image, an artifact boundary and thereby identify the patterned artifact; and
(b) removing the patterned artifact to create a new bitonal image with the patterned artifact substantially removed.
6. The method of claim 5 further comprising the steps of:
(c) identifying text in the new bitonal image; and
(d) cleaning the new bitonal image to remove one or more remaining portions of the patterned artifact and thereby create a second new bitonal image.
7. The method of claim 6 , wherein the step of identifying text in the new bitonal image includes:
identifying text in the new bitonal image;
rank reducing the identified text to create one or more potential blocks of text;
a first removing of one or more remaining portions of the patterned artifact through applying a binary AND operation on the identified text from the identifying text in the new bitonal image step with the one or more potential blocks of text from the rank reducing step to create a first intermediate bitonal image of the new bitonal image;
dilating the first intermediate bitonal image; and
comparing the dilated first intermediate bitonal image with the initial bitonal image within the artifact boundary defined from the identifying the patterned artifact step thereby further removing the one or more remaining portion of the patterned artifact and creating a second intermediate bitonal image of the new bitonal image.
8. The method of claim 7 , wherein the cleaning the new bitonal image step includes:
cleaning the second intermediate bitonal image by applying a binary AND operation on the second intermediate bitonal image with the initial bitonal image within the artifact boundary defined from the identifying the patterned artifact step and thereby create the second new bitonal image.
9. The method of claim 5 , wherein the identifying the patterned artifact step further includes:
before defining the artifact boundary, marking noise locations of the modified image.
10. The method of claim 9 , wherein the removing the patterned artifact step includes:
applying one or more filters in the artifact boundary defined from the identifying the patterned artifact step;
dilating the identified patterned artifact; and
removing areas having a size larger than areas marked as noise from the marking step of the identifying the patterned artifact step and thereby creating the new bitonal image with the patterned artifact substantially removed.
11. A non-transitory computer-readable storage media having stored thereon a plurality of computer-executable instructions for removing a patterned artifact from an initial bitonal image which, when executed by a processor, cause the processor to:
identify the patterned artifact in the initial bitonal image, wherein the identification of the patterned artifact in the bitonal image includes generating a modified image by applying an erosion algorithm based on stroke width of the initial image; and defining, in the modified image, an artifact boundary and thereby identify the patterned artifact;
remove the patterned artifact to create a new bitonal image with the patterned artifact substantially removed; and
clean the new bitonal image to remove one or more remaining portions of the patterned artifact and thereby create a second new bitonal image.
12. The non-transitory computer-readable storage media of claim 11 , wherein the plurality of computer-executable instructions for removing a patterned artifact from an initial bitonal image which, when executed by a processor, further cause the processor to:
prior to cleaning the new bitonal image, identify text in the new bitonal image.
13. The non-transitory computer-readable storage media of claim 12 , wherein the identification of text in the new bitonal image includes:
identifying text in the new bitonal image;
rank reducing the identified text to create one or more potential blocks of text;
a first removing of one or more remaining portions of the patterned artifact through applying a binary AND operation on the identified text from the identifying text in the new bitonal image step with the one or more potential blocks of text from the rank reducing step to create a first intermediate bitonal image of the new bitonal image;
dilating the first intermediate bitonal image; and
comparing the dilated first intermediate bitonal image with the initial bitonal image within the artifact boundary defined from the identifying the patterned artifact step thereby further removing the one or more remaining portion of the patterned artifact and creating a second intermediate bitonal image of the new bitonal image.
14. The non-transitory computer-readable storage media of claim 13 , wherein the cleaning the new bitonal image includes:
cleaning the second intermediate bitonal image by applying a binary AND operation on the second intermediate bitonal image with the initial bitonal image within the artifact boundary defined from the identifying the patterned artifact step and thereby create the second new bitonal image.
15. The non-transitory computer-readable storage media of claim 11 , wherein the removal of the patterned artifact to create a new bitonal image with the patterned artifact substantially removed includes:
applying one or more filters in the artifact boundary;
dilating the identified patterned artifact; and
removing areas having a size larger than areas marked as noise from the marking step of the identifying the patterned artifact step and thereby creating the new bitonal image with the patterned artifact substantially removed.
16. A system comprising:
a processor; and
a computer-readable storage media operably connected to said processor, said computer-readable storage media including instructions that when executed by said processor, cause performance of said processor to remove a patterned artifact from an initial bitonal image by performing operations including:
identify the patterned artifact in said initial bitonal image,
wherein the operation of identify the patterned artifact in said initial bitonal image includes:
generating a modified image by applying an erosion algorithm based on stroke width of the initial image;
marking noise locations of the modified image; and
defining, in the modified image, an artifact boundary and thereby identify the patterned artifact;
remove the patterned artifact to create a new bitonal image with the patterned artifact substantially removed;
identify text in the new bitonal image; and
clean the new bitonal image to remove one or more remaining portion of the patterned artifact and thereby create a second new bitonal image.
17. The system of claim 16 , wherein the operation of remove the patterned artifact to create a new bitonal image with the patterned artifact substantially removed includes:
applying one or more filters in the artifact boundary defined from the identifying the patterned artifact step;
dilating the identified patterned artifact; and
removing areas having a size larger than areas marked as noise from the marking step of the identifying the patterned artifact step and thereby creating the new bitonal image with the patterned artifact substantially removed.
18. The system of claim 16 , wherein the operation of identify text in the new bitonal image includes:
identifying text in the new bitonal image;
rank reducing the identified text to create one or more potential blocks of text;
a first removing of one or more remaining portions of the patterned artifact through applying a binary AND operation on the identified text from the identifying text in the new bitonal image step with the one or more potential blocks of text from the rank reducing step to create a first intermediate bitonal image of the new bitonal image;
dilating the first intermediate bitonal image; and
comparing the dilated first intermediate bitonal image with the initial bitonal image within the artifact boundary defined from the identifying the patterned artifact step thereby further removing the one or more remaining portion of the patterned artifact and creating a second intermediate bitonal image of the new bitonal image.
19. The system of claim 18 , wherein the operation of clean the new bitonal image to remove one or more remaining portion of the patterned artifact and thereby create a second new bitonal image includes:
cleaning the second intermediate bitonal image by applying a binary AND operation on the second intermediate bitonal image with the initial bitonal image within the artifact boundary defined from the identifying the patterned artifact step and thereby create the second new bitonal image.