Back to results
US 2010/0074526 A1Published application
Methods and Systems for Locating Text in a Digital Image
AssigneeSharp Laboratories of America, Inc.
InventorRichard John Campbell
Life of this patent
Estimating the term…
Abstract
Aspects of the present invention are related to systems and methods for locating text in a digital image.
Claims20
A method for locating text in a digital image, said method comprising:
- a) dividing a digital image into a plurality of blocks;
- b) determining a first contrast measure associated with a first block in said plurality of blocks;
- c) determining a first bimodality measure associated with said first block;
- d) classifying said first block as a text block when said first contrast measure and said first bimodality measure meet a first criterion;
- and
- e) classifying said first block as a non-text block when said first contrast measure and said first bimodality measure do not meet said first criterion.
3 more elements
A method locating text in a digital image, said method comprising:
- a) dividing a digital image comprising a plurality of pixels and associated images values into a plurality of blocks;
- b) determining an image-value threshold associated with each block in said plurality of blocks, thereby producing a plurality of image-value thresholds;
- c) determining a contrast measure associated with each block in said plurality of blocks, thereby producing a plurality of contrast measures;
- d) determining a contrast threshold associated with said plurality of contrast measures;
- e) determining a bimodality measure associated with each block in said plurality of blocks, thereby producing a plurality of bimodality measures;
- f) determining a bimodality threshold associated with said plurality of bimodality measures;
- g) determining a bimodal-model-effectiveness parameter associated with said plurality of bimodality measures;
- h) determining an above-threshold measure associated with each block, thereby producing a plurality of above-threshold measures, wherein said above-threshold measure associated with a block is related to the number of pixels in said block with an image value above said image-value threshold associated with said block;
- and
- i) classifying a first block in said plurality of blocks according to said contrast measure associated with said first block when said bimodal-model-effectiveness parameter meets a first criterion;
- and
- j) classifying said first block in said plurality of blocks according to said contrast measure associated with said first block and said bimodality measure associated with said first block when said bimodal-model-effectiveness parameter associated with said plurality of bimodality measures does not meet said first criterion.
9 more elements
Dependencies are inferred from the claim text.