Method for locating bar codes and symbols in an image
First Claim
1. A method for omnidirectional location of bar code symbols and other graphical code symbols in an image of an object, each code symbol having an oriented encoded signal, the method comprising:
- choosing a candidate start region in the image;
determining whether the candidate start region includes at least a portion of a candidate oriented encoded signal, including;
selecting a subset of pixels, from the candidate start region, having enough information for signal detection;
computing a measure of interest of the subset of pixels for use in eliminating candidate start regions that do not contain a signal of interest, without rejecting candidate start regions that might contain a signal of interest; and
deciding whether the measure of interest is sufficient to warrant further examination of the candidate start region;
determining an approximate orientation of the candidate oriented encoded signal; and
growing the candidate start region, using the approximate orientation, so as to form a candidate symbol region that encloses an entire candidate code symbol, including at least a portion of the candidate start region.
3 Assignments
0 Petitions
Accused Products
Abstract
A search method is disclosed for finding bar code symbols and various two-dimensional symbols, regardless of orientation, that is efficient, robust, and easy to use. The method works well on cluttered backgrounds, on skewed symbols, when symbols are almost touching, and when symbols are adjacent to text. The method includes choosing a candidate start region in the image, and then determining whether the candidate start region includes at least a portion of a candidate oriented encoded signal. Next, an approximate orientation of the candidate oriented encoded signal is determined. Then, the candidate start region is grown, using the approximate orientation, so as to form a candidate symbol region that encloses an entire candidate code symbol, including at least a portion of the candidate start region. The method is finely responsive to orientation.
-
Citations
23 Claims
-
1. A method for omnidirectional location of bar code symbols and other graphical code symbols in an image of an object, each code symbol having an oriented encoded signal, the method comprising:
-
choosing a candidate start region in the image; determining whether the candidate start region includes at least a portion of a candidate oriented encoded signal, including; selecting a subset of pixels, from the candidate start region, having enough information for signal detection; computing a measure of interest of the subset of pixels for use in eliminating candidate start regions that do not contain a signal of interest, without rejecting candidate start regions that might contain a signal of interest; and deciding whether the measure of interest is sufficient to warrant further examination of the candidate start region; determining an approximate orientation of the candidate oriented encoded signal; and growing the candidate start region, using the approximate orientation, so as to form a candidate symbol region that encloses an entire candidate code symbol, including at least a portion of the candidate start region. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18)
-
-
19. A method for omnidirectional location of bar code symbols and other graphical code symbols in an image of an object, each code symbol having an oriented encoded signal, the method comprising:
-
choosing a candidate start region in the image; determining whether the candidate start region includes at least a portion of a candidate oriented encoded signal, including analyzing a histogram of edge angles; determining an approximate orientation of the candidate oriented encoded signal; and growing the candidate start region, using the approximate orientation, so as to form a candidate symbol region that encloses an entire candidate code symbol, including at least a portion of the candidate start region.
-
-
20. A method for omnidirectional location of bar code symbols and other graphical code symbols in an image of an object, each code symbol having an oriented encoded signal, the method comprising:
-
choosing a candidate start region in the image; determining whether the candidate start region includes at least a portion of a candidate oriented encoded signal, including; applying an oriented edge detector to the candidate start region to provide a plurality of edge angles; computing an edge angle histogram based on the plurality of edge angles; and analyzing the edge angle histogram to determining whether the candidate start region is likely to include at least a portion of a candidate oriented encoded signal; determining an approximate orientation of the candidate oriented encoded signal; and growing the candidate start region, using the approximate orientation, so as to form a candidate symbol region that encloses an entire candidate code symbol, including at least a portion of the candidate start region. - View Dependent Claims (21)
-
-
22. A method for omnidirectional location of bar code symbols and other graphical code symbols in an image of an object, each code symbol having an oriented encoded signal, the method comprising:
-
choosing a candidate start region in the image, including selecting a rectangular region of the image; determining whether the candidate start region includes at least a portion of a candidate oriented encoded signal, including; selecting several horizontal stripes of the rectangular region as a first subset; computing approximate edge magnitude pixels at a plurality of positions within the several horizontal stripes; computing a first count of edge magnitude pixels by counting how many edge magnitude pixels have an edge magnitude that exceeds a first magnitude threshold; determining whether the first count exceeds a first edge pixel count threshold; selecting at least some of the rectangular region as a second subset; computing edge magnitude pixels at a plurality of positions within the several horizontal stripes; computing a second count of edge magnitude pixels by counting how many edge magnitude pixels have an edge magnitude that exceeds a second magnitude threshold; and determining whether the second count exceeds a second edge pixel count threshold so as to determine whether the candidate start region is interesting; determining an approximate orientation of the candidate oriented encoded signal; and growing the candidate start region, using the approximate orientation, so as to form a candidate symbol region that encloses an entire candidate code symbol, including at least a portion of the candidate start region. - View Dependent Claims (23)
-
Specification