Enhanced Multilayer Compression of Image Files Using OCR Systems

• US 20120294524A1
• Filed: 07/10/2012
• Published: 11/22/2012
• Est. Priority Date: 09/28/2007
• Status: Active Grant

• Current Assignee: ABBYY Development LLC
• Original Assignee: ABBYY Software
• Inventors: Zyuzin, German, Sapronenko, Vyacheslav, Pikhenko, Maksim
• Granted Patent: US 8,548,241 B2
• Time in Patent Office: Days
• US Class Current: 382/166
• CPC Class Codes: G06V 30/413 Classification of content, e.g. text, photographs or tables
• CPC Class Codes: H04N 1/41 Bandwidth or redundancy reduction by scanning H04N1/17 ; methods or arrangements for coding, decoding, compressing or decompressing digital video signals H04N19/00

Abstract

Described herein is a method for segmenting a document image into a picture component, a special or significant picture component, and a non-picture component. The non-picture component is compressed and may include character blocks. Separately, picture components are compressed with a lossy algorithm or with a preliminary defined compression ratio. Subsequently, the compressed picture component, significant picture component and the compressed non-picture component are saved in memory or in a storage location so that the document image may be recomposed based on the compressed picture component or compressed significant picture component and the compressed non-picture component.

Claims

1. A method for processing a document image, the method comprising:

segmenting the document image into a picture component and a non-picture component;

compressing the picture component;

compressing the non-picture component based on a background feature of the non-picture component;

saving the compressed non-picture component and at least one of the compressed picture component in memory; and

forming a recomposed image based on at least the significant picture component and the compressed non-picture component.

2. The method of claim 1, wherein the non-picture component includes characters.

3. The method of claim 2, wherein the method further comprises:

after segmenting the document image, from the non-picture component, constructing a background layer, a chromatic layer and a monochrome mask for the non-picture component.

4. The method of claim 3, wherein the monochrome mask acts as a switch to control when a pixel in the recomposed image would be rendered based on the background and chromatic layers.

5. The method of claim 1, wherein the compressing of the picture component is done based on a preliminary defined compression ratio.

6. The method of claim 1, wherein the compressing of the picture component involves a non-lossy compression algorithm.

7. The method of claim 1, wherein the method further comprises identifying a plurality of character blocks in the non-picture component after segmenting the document image into the picture component, the significant picture component and the non-picture component;

wherein the background layer comprises a background color for each character block in the non-picture component.

8. The method of claim 1, wherein segmenting the document image into the picture component includes identifying each picture in the document image based on one or more chromaticity changes.

9. The method of claim 1, wherein the method further comprises performing optical character recognition on any characters identified in the non-picture component.

10. The method of claim 3, wherein constructing the background layer includes reducing a complexity of the background layer recursively using information about a color of known surrounding pixels to compute a fill color for respective unknown pixels.

11. A system for processing an image of a document, the system comprising:

a processor; and

a memory coupled to the processor, the memory storing instructions which when executed by the processor cause the system to perform a method for processing the image of the document, the method comprising;

segmenting the image of the document into a picture component and a non-picture component;

compressing the picture component;

compressing the non-picture component based on a background feature of the non-picture component;

saving the compressed non-picture component and at least one of the compressed picture component in memory; and

forming a recomposed image based on at least the significant picture component and the compressed non-picture component.

12. The system of claim 11, wherein the non-picture component includes characters.

13. The system of claim 12, wherein the method further comprises:

after segmenting the document image, from the non-picture component, constructing a background layer, a chromatic layer and a monochrome mask for the non-picture component.

14. The system of claim 13, wherein the monochrome mask acts as a switch to control when a pixel in the recomposed image would be rendered based on the background and chromatic layers.

15. The system of claim 11, wherein the compressing of the picture component is done based on a preliminary defined compression ratio.

16. The system of claim 11, wherein the compressing of the picture component involves a non-lossy compression algorithm.

17. The system of claim 11, wherein the method further comprises identifying a plurality of character blocks in the non-picture component after segmenting the document image into the picture component, the significant picture component and the non-picture component;

wherein the background layer comprises a background color for each character block in the non-picture component.

18. The system of claim 11, wherein segmenting the document image into the picture component includes identifying each picture in the document image based on one or more chromaticity changes.

19. The system of claim 11, wherein the method further comprises performing optical character recognition on any characters identified in the non-picture component.

20. The system of claim 13, wherein constructing the background layer includes reducing a complexity of the background layer recursively using information about a color of known surrounding pixels to compute a fill color for respective unknown pixels.