Video stream classifiable symbol isolation method and system
AssigneeKoninklijke Philips Electronics N.V.
InventorsLalitha Agnihotri, Nevenka Dimitrova, Jan Herman Elenbaas
Life of this patent
Estimating the term…
Abstract
An image processing device and method for classifying symbols relies on a connected-component technique for isolating symbol regions. The device and method form connected components from an image derived by the application of an edge detection filter. The formation of connected components from this filtered image defines the edges of character shapes. As a result, the number of pixels that must be connected to define each connected component is substantially reduced and the speed of processing thereby increased. The application of the method is discussed primarily with respect to text in a video stream.
Claims16
A device for classifying symbols in an image data stream containing symbols, comprising:
- an image data storage unit with an input connected to capture data from said image data stream and an output;
- an image processor, connected to said image data storage unit output, programmed to detect edges in portion of captured image data stored in said image data storage unit and form an edge image based on the detected edges, whereby said edge image contains edge pixels and non-edge pixels;
- said image processor being programmed to define at least one connected component from said edge pixels, said connected components being substantially contiguous regions containing only edge pixels;
- said image processor being programmed to define a subimage in said portion whose height and width are substantially the same as a maximum height and width of said at least one connected component;
- and
- said image processor being programmed to classify said subimage, whereby a best match between a symbol set and said subimage is identified.
A method for classifying symbols in an image data stream containing symbols, comprising:
- applying an edge detection filter to an image to identify edge pixels in said image;
- forming connected components from said edge pixels;
- defining symbol-bounding regions each containing a corresponding one of said connected components;
- and
- classifying subimages of said image, each coinciding with a respective one of said symbol-bounding regions.
A method for classifying text in an image data stream containing symbols, comprising:
- applying an edge detection filter to an image to identify edge pixels in said image;
- forming connected components from edge pixels;
- defining text regions based on said forming such that each of said text regions bounds multiple ones of said connected components;
- partitioning said text regions based on spaces between said connected components so as to identify individual character regions of said image;
- and
- classifying subimages of said image, each coinciding with a respective one of said individual character regions.
A device for classifying symbols in an image data stream containing symbols, comprising:
- an image processor connectable to said image data stream to extract an image therefrom and programmed to apply an edge detection filter to said image to identify edge pixels in said image;
- said image processor being further programmed to form connected components from said edge pixels;
- said image processor being further programmed to define symbol regions, each containing a one of said connected components;
- said image processor being further programmed to classify subimages of said image, each coinciding with a respective one of said symbol regions.
An image processor for classifying text in an image data stream containing symbols, comprising:
- an image processing computer connectable to an image data stream to extract an image and programmed to apply an edge detection filter to said image to identify edge pixels in said image;
- said image processing computer being further programmed to form connected components from edge pixels resulting from an application of said edge detection filter;
- said image processing computer being further programmed to define text regions based on said forming such that each of said text regions bounds multiple ones of said connected components;
- said image processing computer being further programmed to partition said text regions based on breaks between said connected components so as to identify individual character regions of said image;
- and
- said image processing computer being further programmed to classify subimages of said image, each coinciding with a respective one of said individual character regions.
Dependencies are inferred from the claim text.