Image analysis – Image segmentation – Distinguishing text from other regions
Reexamination Certificate
1996-12-17
2001-02-27
Tran, Phuoc (Department: 2721)
Image analysis
Image segmentation
Distinguishing text from other regions
C382S155000, C382S161000
Reexamination Certificate
active
06195459
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to the display of images by electronic display apparatus such as “video” type display devices and, in particular, discloses methods and apparatus by which efficient updating of such displays can be obtained.
BACKGROUND ART
Personal computers, computerised work stations and desktop publishing systems, for example, typically incorporate one or more electronic displays by which images are presented to the user to enable manipulation of such systems to perform certain functions and also for the mere presentation of information. Traditionally, such displays have comprised cathode ray tubes devices, which are more commonly known as video displays, and in which image data is displayed in a rasterised format either using interlaced sequential fields or in a non-interlaced manner. Such devices typically have an image display rate of 25 to 30 frames per second (ie. 50 to 60 fields per second), depending upon the particular standard being used. Such arrangements are also utilised for the display of television and other video type signals.
Images displayed on such devices typically comprise text, graphic object animation or symbols, and natural (photographic) type images. Because of the relatively high refresh rate of video displays, there is generally no need to distinguish between different image components in such systems in order to ensure the display of continuous, smooth images. Such arrangements typically require significant computational overhead to ensure regular maintenance of the display and the updating of the image portions thereon.
However, there exist display devices which can operate at lower refresh rates and which permit display management operation utilising lower computational overheads. An example of such a display is a ferroelectric liquid crystal field (FLCD) which can be operated at a refresh rate as low as six frames per second, thus permitting image latency well in excess of that detectable by the human eye. Such low refresh rates permit the computational overhead associated with display management to be reduced compared with that of cathode ray tube devices.
However, difficulties arise with utilising such low refresh for modern computing and desktop publishing applications, for example. In particular, where portions of the image comprise text, it is often the case that those portions of the image are only updated at a relatively slow rate that can be well accommodated by low refresh rates such as six frames per second. However, where complex images such as some animation and natural (photographic-type) images are to be displayed, low refresh rates can provide a significant deterioration in image quality. This is particularly the case with moving images when displayed on such devices, but can also apply to the display of relatively complex static images.
Some methods have been proposed to overcome this difficulty and include halftoning and dithering the image in different manners. However, whilst such methods can provide adequate results for a natural image for example, such systems are not so advantageous for text images, and vice versa. Arrangements have also been proposed which detect movement in portions of the display, utilising threshold techniques which permit identifying that portion of the display that needs to be refreshed at a higher rate. However, halftoning and dithering can give rise to noise in a displayed image that can be misconstrued as movement that can thus cause unnecessary updating of the display and accordingly an overuse of display management systems. These difficulties are also aggravated where the text is coloured on a coloured background, as opposed to the more traditional black on white or white on black text displays.
Accordingly, in order to ensure optimal use of die low refresh rate of such displays, a more efficient means of detecting and distinguishing between text and natural portions of an image is required.
SUMMARY OF THE INVENTION
In accordance with a first aspect of the present invention, there is provided a method of detecting in an image text-like portions and non-text-like portions, said method comprising the steps of:
(a) establishing a set of fuzzy detection rules for distinguishing text-like portions of said image from said non-text-like portions of said image;
(b) dividing the image into a plurality of segments; and
(c) applying said set of fuzzy detection rules to each segment of said image to thereby classify each said segment being one of a text-like portion and a non-text-like portion.
Preferably, the establishing step comprises the sub-steps of:
identifying a plurality of image features that distinguish different portions of an image;
generating a plurality of fuzzy detection rules by applying different combinations of the features to a text-like learning image and to a non-text-like learning image; and
minimizing said plurality of fuzzy detection rules to exclude rules that are not supported by a predetermined amount of the learning images, and allocating the non-excluded rules to the set.
Optionally, the generating sub-step comprises the sub-steps of:
normalising each image feature as a vector to have a value in the range 0 to 1;
partitioning each input feature space into a plurality of equally spaced regions;
assigning each input feature to a label of one of the regions to maximize a membership value of the label in the one region;
selecting for each the region the maximized label for each the feature to thus form a respective fuzzy rule.
In accordance with a second aspect of the invention, there is provided an apparatus for detecting in an image text-like portions and non-text-like portions, said apparatus comprising:
means for establishing a set of fuzzy detection rules for distinguishing text-like portions of an image from non-text-like portions of an image;
means for dividing the test image into a plurality of segments; and
means for applying said set of fuzzy detection rules to each segment of said test image to thereby classify each said segment as being one of a text-like portion and a non-text-like portion.
In accordance with a third aspect of the invention, there is provided a method of zone segmenting a digital image for display on display means, wherein said digital image is processed as a plurality of blocks each having a predetermined number of pixels, said method comprising the steps of:
extracting a set of features from each block to generate a feature vector for said block;
classifying said block using a set of fuzzy rules as either a text-type image or a natural-type image dependent on said feature vector for said block.
In accordance with a fourth aspect of the invention, there is provided an apparatus for zone segmenting a digital image for display on display means, wherein said digital image is processed as a plurality of blocks each having a predetermined number of pixels, said apparatus comprising the steps of:
means for extracting a set of features from each block to generate a feature vector for said block;
means for classifying said block using a set of fuzzy rules as either a text-type image or a natural-type image dependent on said feature vector for said block.
In accordance with a fifth aspect of the invention, there is provided a display method for generating a display signal for display on a display device in response to an input video signal, said method comprising the steps of:
partitioning a frame of said input video signal into a plurality of blocks, where each block comprises a predetermined number of pixels;
extracting a set of features from each block to generate a feature vector for said block;
identifying each block using fuzzy rules as either a text-like image or a natural-like image dependent on said feature vector for said block;
storing a plurality of identified blocks in a memory;
generating said display signal at a lower data rate than said input video signal dependent upon said identified blocks, wherein a first region of said frame comprising a number of identified text-like blocks are generated differently fro
Canon Kabushiki Kaisha
Fitzpatrick ,Cella, Harper & Scinto
Tran Phuoc
LandOfFree
Zone segmentation for image display does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Zone segmentation for image display, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Zone segmentation for image display will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2607684