AU720754B2 - A method of smoothing bitmapped characters - Google Patents

Description

wP, S F Ref: 408029

AUSTRALIA

PATENTS ACT 1990 COMPLETE SPECIFICATION FOR A STANDARD PATENT

ORIGINAL

r r Name and Address of Applicant: Actual Inventor(s): Address for Service: Invention Title: Canon Kabushiki Kalsha 30-2, Shimomaruko 3-chome Ohta-ku Tokyo 146

JAPAN

Jon Windle Spruson Ferguson, Patent Attorneys Level 33 St Martins Tower, 31 Market Street Sydney, New South Wales, 2000, Australia A Method of Smoothing Bitmapped Characters ASSOCIATED PROVISIONAL APPLICATION DETAILS [31] Application No(s) [33] Country P05381 AU [32] Application Date 28 February 1997 The following statement is a full description of this invention, including the best method of performing it known to me/us:- 5815 -1- A METHOD OF SMOOTHING BITMAPPED CHARACTERS Field of the Invention The present invention relates to the display of bitmapped data and, in particular, to the display of bitmapped text on a display device that has a greater resolution than the intended resolution of the text.

Background Art The clarity of text is closely linked to the resolution of a display device. At higher resolutions, text can be made smoother and easier to read. Greater resolution also enables smaller text to be read more easily.

A device able to be connected to a computer with a minimum (or no) amount of user time required for configuring the device to the computer, is often referred to as a "Plug and Play" device. For example a plug and play device can be an ferro-electric liquid crystal display (FLCD) pre-configured to interface with a plurality of computer systems with a minimum of user interaction.

Typically, an FLCD panel is capable of displaying text and natural image data, at a standard screen resolution of 89 pixels per inch with a colour range of 4096 distinct colours. Such a panel is also capable of double the resolution with a limited colour range. However, to display text data, at the double resolution, additional information is required of the text data to take advantage of the higher resolution the FLCD panel.

Current text enhancement systems rely on characteristics of a particular device.

For example, in laser printers the area of a recorded dot can be controlled by controlling the characteristics of the laser. Other systems, for text enhancement relies Sg on storing a plurality of pixel lines, generally at least three, of the text data in order to identify a pixel bit pattern. For coloured text and background, these techniques require 25 storing pixel values, which typically include three colour components such as Red, oO.o *o**Green and Blue (RGB) for each pixel. With 8 bits for each colour component, 24 bits/pixel are required to be stored.

The foregoing presents a number of disadvantages including device dependence and the requirements to store pixel colour value information. Further disadvantages include a complexity of extracting information from the design font of text characters.

It is therefore desirable to provide a method of generating additional data that lends assistance to improving the overall appearance text data, that can be implemented with less memory than existing systems, and is substantially colour independent. A solution would preferably have the characteristic of minimal (or no) user interaction, S 35 improved overall appearance of the text, even if individual letters are not appreciably improved, and the ability to use information implicit in a font design of the text data.

Summary of the Invention It is an object of the present invention to substantially overcome, or ameliorate one or more problems associated with existing arrangements.

FLCD19 408029 CFP0757AU [o:\cisra\flcdncw%\flcd 19]408029.doc In accordance with a first aspect of the present invention there is provided a method of extracting information from a bitmap image representation, the bitmap image representation comprising at least a portion of a bitmapped character formed from a plurality of pixels, said information being inherent in the arrangement of pixels of the bitmapped character, said method comprising the steps of: generating a modified data representation for the bitmap image representation wherein said modified data representation is characterised by additional information being inherent in the arrangement of pixels in the bitmap character; and identifying one or more predetermined features from said modified data representation.

In accordance with a second aspect of the present invention there is provided a method of enhancing a bitmapped character, said bitmapped character comprising a plurality ofpixels, said method comprising the steps of: inputting a bitmap image representation including at least a portion of the bitmapped character; generating a modified data representation for each pixel of the bitmap image representation wherein said modified data representation is characterised by additional information being inherent in the arrangement of pixels in the bitmap character and storing said modified data representation; identifying one or more predetermined features relating to said bitmapped character from the stored modified data; and generating an output bitmap image representation based on the modified data representation and the identified predetermined features.

In accordance with a third aspect of the present invention there is provided a 25 bitmapped character enhancing apparatus comprising; input means for inputting a bitmap image representation comprising at least part of a.

a bitmapped character; data modification means connected to said input means for providing a modified 3 data representation wherein said modified data representation is characterised by additional information being inherent in the arrangement of pixels in the bitmap character; data storage means, coupled said data modification means, for storing said modified data representation; feature extraction means for extracting from the modified data representation predetermined features of said bitmapped character; and data generation means for providing an output bitmap image representation based a. S-on the modified data representation and the predetermined features; In accordance with a fourth aspect of the present invention there is provided a R bitmapped character enhancing apparatus comprising: 408029 CFP0757AU [o:\cisra\flcdnew\flcdl9]408029.do input means for inputting a bitmap image representation comprising at least part of a bitmapped character; data modification means connected to said input means for providing a modified data representation of said bitmapped character wherein said modified data representation is characterised by additional information being inherent in the arrangement of pixels in the bitmap character; data storage means, coupled said data modification means, for storing said modified data representation; feature extraction means for extracting from the modified data representation of predetermined features of said bitmapped character; data generation means for providing a corrected bitmap image representation of said bitmapped character based on the modified data representation and the extracted predetermined features; pixel data storage means, coupled to the input means for receiving and storing said input bitmap image representation; and data combining means for receiving the input bitmap image representation and corrected bitmap image representation, from the pixel data storage means and said data generation means respectively, and for producing an output bitmap image representation based on the input bitmap image representation and corrected bitmap image representation.

In accordance with a fifth aspect of the present invention there is provided a method of displaying a bitmapped character formed from a bitmapped image representation including a plurality of pixels intended for display at a first display resolution, said method comprising the steps of: 1" 25 generating a modified data representation for the bitmapped image representation, said modified data representation comprising a plurality of pixels at a second display resolution for each pixel at the first display resolution and wherein said modified data representation is characterised by additional information being inherent in 0 the arrangement ofpixels in the bitmap character; identifying from said modified data representation at least one predetermined *fewfeature of said bitmapped character being inherent in the arrangement of pixels; ooo0 altering said modified data representation in accordance with a predetermined feature; creating an altered bitmap image representation of said bitmapped character based on the altered modified data representation; and displaying the altered bitmap image representation with a display device at the second display resolution.

[o:\cisra\flcdnew\flcdI 9]408029.doc -3A- Brief Description of the Drawings Embodiments of the present invention will now be described with reference to the Fig. 1 shows an example of a text characters, a letter displayed as a pixel based image at a first resolution; Fig. 2 illustrates the character of Fig. 1 displayed at a second (higher) resolution; Fig. 3 is a representation of a pixel, at the first resolution, divided into four qudans U' LD19 408029 CFP0757AUJ [o:\cisra\flcdniew\flcdl9]408029.doc sq 5*

S

S S

S

*5 S. 55 S

S

S.

5

S

S.

S

S

9

S

@55e

S

9 S. *5 S.

S

S

-4- Fig. 4 illustrates the character of Fig. 1 displayed at the second (higher) resolution and smoothed in accordance with an embodiment of the present invention; Fig. 5 is a schematic block diagram representation of an apparatus in accordance with a first embodiment of the present invention; Fig. 6 is a representation of a single pixel at foreground colour (black) surrounded by a plurality of background colour pixels (white); Fig. 7 illustrates a modified data representation of the pixels of Fig. 6; Fig. 8 is an illustrated representation of a single pixel in background colour (white) surrounded by at least four foreground coloured pixels (black), referred to herein as a "pixel hole"; Fig. 9 illustrates an "uncorrected output" for the pixel hole of Fig. 9; Fig. 10 is an illustration of a "corrected output" corresponding to the "uncorrected output" of Fig. 9 in accordance with the embodiment-of the present invention; Fig. 11, Fig. 12 and Fig. 13 illustrate several examples of features to be detected in accordance with one or more embodiments of the present invention; Fig. 14 illustrates various stages of processing to smooth text character in accordance with a first embodiment; Fig. 15A to Fig. 20B illustrate various features to be detected, a corresponding corrected output and a corrected output are also shown for each feature in accordance with embodiments of the present invention; wtFig. 21 is a schematic block diagram representation of an apparatus in accordance with a record embodiment; Fig. 22 is an alternate modified data representation in accordance with a third S* 25 embodiment; and Fig. 23 is an illustrated example of smoothing at various (higher) resolutions in accordance with any one of the embodiments of the present invention.

Unless otherwise specified, throughout a detailed description of the embodiments, reference numerals corresponding to substantially similar integers of the accompanying drawings have the same value.

Detailed Description Video information, including computer generated information, typically comprises data information for individual picture elements (pixels). For example, in Fig. 1, a text character 100 is shown displayed on a discrete display device and made 35 up of a series of pixels, in this case black pixels 101 representing a foreground colour and white pixels 102 representing a background colour which combined together to form the letter The data information is generated and intended for display on a display device at a first resolution (hereinafter referred to as "normal" resolution) and displaying the same information on a display device having a second (higher) resolution FLCD19 408029 CFP0757AU [o:\cisra\flcd new\llcd 19]408029.doc (hereinafter referred to as the higher resolution) does not necessarily take advantage of the increased resolution. Fig. 2 illustrates the text character 100 of Fig. 1 displayed at a second (higher) resolution and scaled such that the text character has substantially the same physical dimensions (size) as the character of Fig. 1. As seen from Fig. 2, by displaying the text character on a display device having a higher resolution, and scaling the text character to similar dimensions as the intended size of the text character, does not necessarily take advantage or substantially rely upon the increased resolution of the display device. A pixel 201 of Fig. 2 is a quarter the size (dimensionally), of a pixel 102 of Fig. 1, and therefore four times the number of pixels are required to reproduce substantially the same character as that shown in Fig. 1. Although Fig. 2 provides an increased resolution advantage, the character 100 itself has not been modified to advantageously make use of the increased resolution.

Smooth Processing System A First Embodiment The first embodiment of the present invention will be described with reference to a Ferroelectric Liquid Crystal Display (FLCD) capable of a normal screen resolution display of 89 pixels per inch with a colour range of 4096 colours and also capable of higher resolution by utilising sub-pixels of the display device. However, it is envisaged that the present invention can be utilised for text characters intended for display (or displayed) at a first resolution on a first display device, and displayed at a second (higher) resolution on a second display device. An example of this includes text characters displayed on a typical computer monitor at first resolution to be printed on a print device having a greater resolution. Another example includes characters intended for display on a low-resolution display device (eg. low resolution computer monitor), :°°°.which are displayed on a high-resolution display device (eg. high resolution computer monitor).

Further, while the present embodiments are described herein with reference to black pixels representing the foreground colour and white pixels representing the background colour, this is only to simplify the present description, and any colour combination can be chosen representing the foreground and background colours, other *°°00 than the foreground colour being the same as the background colour.

Turning to Fig. 3, there is shown the pixel 102 of Fig. 1 divided into a plurality of quadrants 310 to 313, with each quadrant representing a pixel 201 (or sub-pixel of an FLCD display) of the increased resolution display of Fig. 2. To simplify the o description of the embodiments, a reference to the term "quadrant" is to be construed, 35 unless otherwise noted, as representing a pixel at the second (high) resolution. Further, an example is shown in Fig. 3 of a labelling of quadrants and, unless otherwise noted, this labelling is used throughout the present specification. In particular the pixel 102 is divided into a top left quadrant 310, a top right quadrant 311, a bottom left quadrant 312 and a bottom right quadrant 313.

FLCD19 408029 CFP0757AU [o:\cisra\flcdncw\flcd I 91408029.doc Referring now to Fig. 4, there is shown the text character of Fig. 1 displayed at the second (higher) resolution and smoothed in accordance with the first embodiment of the present invention. The computer implemented process of smoothing the text character is determined from information that is implied in the font design of the text character itself. For example, in Fig. 4 the relative positions of pixel 203 and pixel 204 indicates that the text character would benefit in the overall appearance if quadrant 401 and quadrant 402 were switched to substantially the same colour as the foreground colour. In this example, the quadrants 401, 402 would thus be switched from the colour white to the colour black.

Turning to Fig. 5, generated information 498 representing both text data information and natural image information is input to a zone segmentation unit 499.

The text data information includes text characters (including oriental characters as distinct from occidental characters) and line drawings. Natural image information includes, within its scope, images such as computer generated images or images electronically scanned via a scanner other than text data. The zone segmentation unit 499 discriminates between text data information and natural image information and outputs the text data information separately. The text data information from the zone segmentation unit 499 is input to a smoothing processing system 500 via an input 501.

A data modification unit 502, connected to the input 501, modifies the incoming text data information. The manner in which the text data information is modified will be described hereinafter under the heading "Data Modification". The data modification unit 502 outputs the modified data information to a memory storage means 503 which stores at least a portion of the modified data information. Preferably, the memory storage means 503 is random access memory (RAM) capable of storing the equivalent of two successive pixel lines of the text data in a modified data representation. A feature extraction unit 504 connected to the memory storage means 503 extracts from the modified data information, features or attributes of the modified data information that enables a data generation unit 505 to determine the required alterations to the modified data information so as to produce for output a smoothing of text characters or lines represented by the text data information. The modified data information is passed to the feature extraction unit 504, from the memory storage means 503, and then to the data generation unit 505 together with the features or attributes extracted by the feature g. extraction unit 504.

The output of the data generation unit 505 is smoothed text data information and is directed to a display device (not shown in Fig. 5) via an output 506. An example of smoothed text data information in accordance with the first embodiment is illustrated by Fig. 4.

Preferably the text data information is supplied to the smoothing processing system 500 on a pixel by pixel basis in raster image order. For the purpose of FLCDI9 408029 CFP0757AU [:\cisra\lcdniew nc 19]408029. doc -7simplifying the description of each embodiment, herein described, the raster order is assumed to be from left to right across a display device and from top to bottom of the display device in a non-interlaced fashion. Typically, the foreground colour and/or the background colour of a text character, represented by the text data information, is supplied to the smoothing processing system 500 by the zone segmentation unit 499, however it is sufficient that the foreground colour, background colour or a manner of determining both or either of these colours is available to the smoothing processing system 500.

Data Modification The data modification unit 502 accepts text data information on a pixel by pixel basis as previously described and stores in the memory storage means 503 a modified data representation in accordance with the following rules: 1) if a pixel input to the data modification unit 502 is a background colour, store in memory 503 four (zeros), one for each quadrant representing the input pixel; 2) if a pixel input to the data modification unit 502 is a foreground colour then: i) add (one) to all the quadrants in memory 503 that represent this pixel; and ii) add "0.25" (a quarter) to the bottom quadrants of a pixel above and to the top quadrants of a pixel below the current input pixel; 3) if a previous pixel (a pixel to the left of the current pixel) is a background colour, add "0.25" (a quarter) to the right quadrants of the previous pixel in memory 503; if a next pixel (a pixel to the right of the current pixel) is a background 25 coloured pixel add "0.25" to the right quadrants that represent this pixel; and 5) repeat the above for each input pixel.

Referring to Fig. 6, there is shown a schematic example of a portion 701 of text data information having one pixel 702 of foreground colour and surrounded by a plurality of pixels 703 of background colour. An exemplary pseudo-code procedure for processing a pixel into a modified data representation is provided below and the procedure is referred to as "ProcessPixelLine".

PROC ProcessPixelLine FOR EachPixellnThelnputLine IF CurrentPixelColour EQUALS ForegroundColour THEN add 1.0 to all quadrants for the current pixel AddPixelVal(Curr, PixelNumber, ALL, FLCDI9 408029 CFP0757AU [o:\cisra\lcdlew\flcd 191408029.doc add 0.25 to the bottom quadrants of the pixel above this one, i.e. same pixel number previous line.

AddPixelVal(Prev, PixelNumber, BOTTOM, 0.25) add 0.25 to the top quadrants of the pixel below this one, i.e. same pixel number next line.

AddPixelVal(Next, PixelNumber, TOP, 0.25) check for edge conditions IF NOT LastPixellnLine THEN IF NOT FirstPixellnLine THEN IF NextPixelCol EQUALS BackgroundCol THEN add 0.25 to the left quadrants of the pixel after this one.

AddPixelVal(Next, PixelNumber 1, LEFT, 0.25)

ENDIF

IF PreviousPixelCol EQUALS BackgroundCol THEN add 0.25 to the right quadrants of the pixel before this one.

o AddPixelVal(Next, PixelNumber-1, RIGHT, 0.25)

ENDIF

ENDIF

ENDIF

END IF GET NEXT PIXEL END PROC Fig. 7 illustrates a modified data representation 810, corresponding to the portion 701 of text data information shown in Fig. 6. Four quadrants 811 having a value of (one) correspond to the pixel 702 of foreground colour, and quadrants 812 immediately above quadrants 813 immediately below, quadrants 814 to the immediate left and quadrants 815 to the immediate right of the four quadrants 811 each take on a value of "0.25" (a quarter) in accordance with the modification rules set out above.

Referring again to Fig. 5, the feature extraction unit 504 receives a modified data representation from the memory storage means 503 and for each current quadrant FLCD19 408029 CFP0757AU [o:\cisra\llcd,,ew\flcd] I9]408029.doc -9having a modified data representation value greater than or equal to a predetermined value (eg. the current quadrant is marked as having a foreground colour. For each current quadrant having a modified data representation value below the predetermined value, the current quadrant is marked as having a background colour.

To simplify the description of the function of the feature extraction unit 504, the quadrants are taken as having been assigned their respective foreground or background colours rather than simply being marked as having a foreground colour or background colour as described above. The modified data information represented as foreground and background colour quadrants is referred to hereinafter as an "uncorrected output" and is an output of the feature extraction unit 504 directed to the data generation unit 505. Thus, in the uncorrected output, each quadrant with a modified data representation equal to, or greater than, a predetermined value (eg. is assigned a foreground colour (black), while each quadrant having a value below the- predetermined value is assigned a background colour (white).

In addition to determining which quadrants are of background colour and foreground colour, the feature extraction unit 504 acts to identify from the modified data representation predetermined features (characteristic attributes) which indicate a predetermined positional relationship between a plurality of pixels.

Based on the modified data representation, the feature extraction unit 504 determines whether a quadrant has characteristic attributes which belong to a set of predetermined features. An example of a set of predetermined features will be described herein under the heading "Features to Extract". This characteristic attribute o information is passed to the data generation unit 505 along with the "uncorrected, output" previously described. The data generation unit 505 determines what further S• 25 correction is required to the uncorrected output before display on the display device.

The output of the data generation unit is directed to the display device, and the output of the data correction unit is referred to herein as the "corrected output". If the feature 0 to extraction unit 504 cannot determine a characteristic attribute for a quadrant, or the characteristic attribute cannot be identified from a set of predetermined features, then a signal from the feature extraction unit 504 to the data generation unit 505 notifies the data generation unit 505 that no further correction to the "uncorrected output" is to be made before display. In other words, the "uncorrected output" is output by the data generation unit 505 as if it were "corrected output" when none of the predetermined features are identified.

to.0 35 The characteristic attribute information identified (or determined) from the t a modified data representation includes determining for a current pixel a total quadrant value being a sum of all quadrant values for the current pixel, the total quadrant value is used as an indicator, of a particular feature, from a set of predetermined features.

Features To Extract FLCD19 408029 CFP0757AU [o:\cisra\tlcdinew\flcd9148O29.doc Without the feature extraction unit identifying characteristic attributes and the data generation unit further correcting the uncorrected output, the resulting display of text can have an overall "lumpy" or "blobby" appearance. The feature extraction unit 504 identifying at least some features, and the data generation unit 505 correcting the uncorrected output accordingly has the effect of removing the "lumps"/"blobs", giving the displayed text a sharp clean appearance.

1) Single Pixel Hole With reference to Fig. 8, a single pixel hole 900 can be described as a single pixel 901 having a background colour, substantially surrounded by a plurality of pixels 902 having a foreground colour. Further, a key is provided in Fig. 8 indicating the convention used in this description for each pixel colour, where a black coloured pixel represents a pixel having foreground colour, a white coloured pixel represents a pixel having a background colour, and a shaded pixel marked with cross-hatches can be optionally either foreground or background colour. This convention is adopted in Fig.

8 through to Fig. In text data information comprising small fonts, single pixel holes are readily encountered, for example, in the dot of a character in predetermined fonts. A single pixel hole can be identified from the modified data representation previously described, by checking for a pixel having a total quadrant value of 2.0. In the present example, a total quadrant value of 2.0, for a current pixel, indicates that the pixel is surrounded by at least four foreground pixels, one on each side.

A representation of an uncorrected output 1003, from the feature extraction unit •504, for a single pixel hole 1000 is shown in Fig. 9. A corresponding corrected output 1100 for the single pixel hole 1000 of Fig. 9 is illustrated in Fig. 10 and is a S 25 representation of the output resulting from the data generation unit 505.

To correct the uncorrected output 1003, for a single pixel hole, the data generation unit 505 re-assigns to each quadrant of a centre pixel 1004 of the uncorrected output, a background colour. In general, a pixel that has a background colour on input to the smoothing processing system 500, can have a maximum total 09* quadrant value of 2.0 when the background pixel is surrounded by a foreground pixel on each side; above, below, to the left and to the right of the background pixel. A pixel that has a foreground colour on input to the smoothing processing system 500, 9 can have a maximum total quadrant value of Consequently a pixel found to have a total quadrant value of 2.0, in the modified data representation, is indicative of a pixel hole and therefore is corrected by assigning a value of zero to each quadrant of the pixel.

The detection (identification) and correction of single pixel holes enhances the visual appearance of the single pixel hole when compared with the uncorrected output quality.

FLCD19 408029 CFP0757AU [o:\cisra\flcdew\flcd I 91408029.doc -11- 2) Pixel Wide Gaps Referring to Figs. 11 to 14 there are illustrated some examples of pixel wide gaps. A common feature of text characters is the occurrence of pixel wide gaps and these gaps can present themselves in either a horizontal or vertical orientation. A gap is characterized as a "closing gap" in a vertical orientation if background pixels (white pixels) decrease in number progressively down the gap from top to bottom.

Consequently, a gap is said to be an "opening gap" in a vertical orientation if background pixels increase in number progressively down the gap from top to bottom.

Similarly, gaps are described as opening or closing in a horizontal orientation if background pixels increase or decrease in number respectively, progressing from left to right along the gap. For example the vertex of the character is one type "closing gap" indicated at 1201 in Fig. 11 having a vertical orientation and typically closing to substantially one pixel at its vertex. Another example comprising a plutrality of pixel wide gaps is the letter which includes a vertical stroke and an upper slanted stroke forming a vertically oriented closing gap 1202 (Fig. 11), a horizontally oriented opening gap 1401 (Fig. 13) is found between the upper slanted stroke and a lower slanted stroke, and a vertically oriented opening gap 1402 (Fig. 13) is found between the vertical stroke and the lower slanted stroke.

In Fig. 12, there are two examples of constant size single pixel wide gap (1301, 1302) such constant size single pixel wide gaps are generally encountered in small font text and in particular characters like in small font.

A total quadrant value of 1.5, from the modified data representation, is sufficient to identify an end pixel of a pixel wide gap. For example, in a vertically oriented So°°° closing gap (1201, 1202), an end pixel (1203, 1204) is a pixel at the narrow end of the V,004 25 gap. As a further example, in a constant size single pixel gap (1301,1302), an end o pixel (1303,1304) is a background pixel bounded on three sides by a foreground pixel.

g It Detecting a total quadrant value of 1.5, however, is not generally sufficient to determine the orientation (vertical or horizontal) or type (opening or closing) of a gap.

More information, from the modified data representation, of at least one surrounding o*•SS 30 quadrant is required.

Firstly, for a current pixel having a total quadrant value of 1.5, a gap orientation *is determined by checking for a predetermined ordering of quadrant values for the s'e current pixel. Using the notation developed for Fig. 3 and describing an end pixel as having quadrants arranged in order from top left 310, top right 311, bottom left 312 to S 35 bottom right 313, the following gap orientations can be determined:

S.=

an end pixel whose quadrant values are 0.25, 0.25, 0.5, 0.5 respectively is the end pixel of a vertically oriented closing gap or a constant size single pixel gap with a gap opening facing upward; FLCD19 408029 CFP0757AU [o:\cisra\flcdncw\flcd 19]408029.doc -12an end pixel whose quadrant values are 0.5, 0.5, 0.25, 0.25 respectively is the end pixel of a vertically oriented opening gap or a constant size single pixel gap with a gap opening facing downward; a an end pixel whose quadrant values are 0.25, 0.5, 0.25, 0.5 respectively is the end pixel of a horizontally oriented closing gap or a constant size single pixel gap with a -gap opening facing to the left; and an end pixel whose quadrant values are 0.5, 0.25, 0.5, 0.25 respectively is the end pixel of a horizontally oriented opening gap or a constant size single pixel gap with a gap opening facing to the right.

Once the gap orientation has been determined, each quadrant of the end pixel having a quadrant value 0.5 or greater marked as having a foreground colour) is further analysed to determine if the quadrant is to remain marked as foreground or switched to a background colour.

The further analysis of the end pixel in a pixel wide gap is described with reference to Fig. 14. For the sake of simplicity, only a single example is described in detail. However, other examples shown in Fig. 15B can be described in a substantially similar manner. In the present example, the input character is a small (three pixel wide) letter 1500, and a corresponding modified data representation 1501 for the character 1500 is shown and is determined in accordance with the data modification 20 rules previously described. At the uncorrected output stage 1512, that is the output of the feature extraction unit 504, each quadrant value greater than or equal to 0.5 (half) is assigned a foreground colour (or mark accordingly).

The uncorrected output 1512 for the letter 1500, if it were to be displayed on i. a display device, would take on a smooth appearance 1502 but a gap 1503 between vertical strokes of the character would decrease in size when compared to a corresponding gap 1513 of the input character 1500.

The feature extraction unit 504 determines the orientation of the gap 1503 from the representation data 1501 as previously described and further analysis of the modified representation data is performed on quadrants having a modified data value equal to The further analysis includes applying one, or both, of two tests set out below.

i •First Test For a left quadrant 1504 of a current pixel 1505, having a value of 0.5 (half), a check is performed on a first quadrant 1506 to the immediate left of the left quadrant 1504 of the current pixel 1505. If the value of the first quadrant 1506 is 1.0 (one) then the left quadrant 1504 of the current pixel is to retain the value of 0.5 (half) and the first test is applied to a next quadrant whose value is 0.5 (half). However, if the value of the first quadrant 1506 is not 1.0 (one) a second test is to be applied. The second test is described hereinafter.

FLCD19 408029 CFP0757AU [o:\cisra\flcdnnew\ fcd j9]408029.doc -13- For a right quadrant 1507 of a current pixel 1505, having a value 0.5 (half), a check is performed of a second quadrant 1508 to the immediate right of the right quadrant 1507 of the current pixel 1505. If the value of the second quadrant 1508 is (one) then the right quadrant of the current pixel 1504 is to retain the value (half) and the first test is applied to a next quadrant whose value is 0.5 (half).

However, if the value of the second quadrant 1508 is not 1.0 (one), the second test is to be applied.

In the present example, illustrated in Fig. 14, the right quadrant 1507 has an adjacent quadrant 1508, to the immediate right, with a value of 1.0 hence the left 1o quadrant 1507 retains a value 0.5 (half). The left quadrant 1504 has an adjacent quadrant 1506, to the immediate left with a value of 1.25, therefore a second test must be applied to the left quadrant 1504 to determine if the value of 0.5 is to be retained or a new value is to be assigned to the quadrant 1504.

Second test Preferably, the second test is applied as a consequence of the first test and where the first test cannot determine if a quadrant, having a value of 0.5 (half), is to retain its value as previously described. The second test need not be applied where the first test successfully determines that each quadrant, having a value of 0.5 (half), can retain that value.

20 The second test for a pixel wide gap is set out below and again for the purpose of simplifying the description, this test will also be described with reference to Fig. 14.

Unless otherwise noted, a reference to a gap's "forward direction" is to be construed, hereinafter, as a reference to a direction pointing from a closed end of the gap to an open end of the gap. For example, the forward direction of the gap 1509 of the character is shown by an arrow 1510 pointing upward out of the "mouth" or *.open end of the gap 1509. The second test determines if a pixel diagonally adjacent in a forward direction to a current pixel (a current pixel having at least one quadrant of value 0.5) is of foreground or background colour. In particular, for the left quadrant 1504 having a value of 0.5, a check is made to determine if a left diagonally adjacent pixel 1511 in the forward direction is of foreground or background colour.

If the right diagonally adjacent pixel 1511 is of foreground colour, as seen for example in Fig. 14, then the left quadrant 1504 of the current pixel is re-assigned a o value of zero.

Otherwise, the right diagonally adjacent pixel 1511 is of background colour and therefore the value of 0.5 is to be retained by the right quadrant 1504 of the current pixel 1505.

An application of the second test to the left quadrant 1507 is described substantially as for the right quadrant 1504, excepting that for the left quadrant 1507 of the current pixel 1505, colour information of a left diagonally adjacent pixel 1514 is FLCDI9 408029 CFP0757AU \cisra\flcdnecw\flcd 19l408029. doc 14used to determine whether or not the left quadrant 1507 is to retain a value of However, in the present example, the left quadrant is to retain the value of 0.5 as determined by the first test described above. Consequently, the second test is not to be applied to the left quadrant 1507 of the current pixel 1505.

Referring now to Fig. 15A, an example of a pixel wide constant gap 1600 is shown as input data to the smooth processing system and a corresponding uncorrected output 1601 representation resulting from the feature extraction unit. In this example a black pixel represents a foreground colour, a white pixel a background colour and a shaded or cross-hatched pixel represents optionally either background or foreground colour. Therefore, by judiciously choosing the shaded/cross-hatched pixels to be either background or foreground, at least five case options for a vertically oriented pixel wide gap can result. Consequently, Fig. 15A represents the input and uncorrected output for at least five cases.

Turning to Fig. 15B, there is shown five exemplary cases of input pixel wide constant gaps 1602 and their corresponding corrected output 1603 for each input. In each case the corrected output 1603 is a desired modification to be performed to the uncorrected output of Fig. 15A when smoothing the pixel wide constant gap in accordance with an embodiment of the present invention. The desired modification to the uncorrected output 1601 may include, as for Case 4 example of Fig. 15B, no i 20 modification to the uncorrected output 1601.

For example, if a character, or part of a character, input to the smoothing processing system 500 consists of a pixel wide constant gap 1602, shown in Case 4 of Fig. 15B, then the corresponding corrected output 1603 takes on a form substantially similar to the corresponding uncorrected output 1601.

Fig. 15B represents a plurality of vertically oriented examples of constant pixel wide gaps 1602 and their corresponding corrected output 1603. Similarly, horizontally oriented constant pixel wide gaps, both orientation of closing gaps such as those depicted in Fig. 11 and both orientation of opening gaps (examples shown in Fig. 13) have associated a plurality of cases for each orientation and type of gap opening, closing, constant).

The pixel wide gap detection and correction does not yield desired results in all pixel wide gaps for all characters, however a noticeable difference in the overall smoothing effect is visible at least for text characters such as and There are a number of cases where a corner (or edge) of a character can be "thickened", however this does not seem to noticeably degrade the output quality of the overall appearance of text characters. In particular, Case 3 and Case 4 of Fig. 15B show the corrected output 1603 have a shorter gap when compared to the input pixel wide gap 1602, hence a noticeable thickening of lower edge of the input pixel wide gap. An example of a pseudo-code for determining whether or not a pixel wide gap has a FLCD19 408029 CFP0757AU [o:\cisra\1cdnew\1cd 191408029.doc vertical gap orientation is provided below. Pseudo-code for determining whether, or not, a pixel wide gap has a horizontal orientation has not been included, however is substantially similar to the pseudo-code for determining the vertical orientation set out below.

IF QuadrentSum EQUALS 1.5 THEN it is the end of a pixel gap so determine it's orientation TopLeftQuad CurrentPixel.TopLeftQuadrent IF ((PreviousPixel EQUALS ForegroundColour) AND (NextPixel EQUALS ForegroundColour)) THEN Gap is vertical IF TopLeftQuad EQUALS 0.5 THEN GapDirection DOWNWARD, check to see if the quadrants Should be switched to the background colour IF (NextPixel.TopLeftQuadrant GREATER THAN 1.0) THEN pixel forward and up is a foreground one IF (NextPixel.BottomLeftQuadrant GREATER THAN 1.0) THEN 20 pixel forward and down is a foreground one CurrentPixel.TopRightQuadrant 0

ENDIF

IF (PreviousPixel.TopRightQuadrant GREATER THAN 1.0) THEN pixel backward and up is a foreground one IF (PreviousPixel.BottomLeftQuadrant GREATER THAN

THEN

pixel backward and down is a foreground one CurrentPixel.TopLeftQuadrant 0

ENDIF

30 ENDIF

ELSE

GapDirection UPWARD IF (NextPixel. BottomLeftQuadrant GREATER THAN 1.0) THEN pixel forward and down is a foreground one IF (NextPixel.TopLeftQuadrant GREATER THAN 1.0) THEN pixel forward and up is a foreground one CurrentPixel. BottomRightQuadrant 0

ENDIF

IF (PreviousPixel.BottomRightQuadrant GREATER THAN 1.0) THEN FLCD19 408029 CFP0757AU [o:\cisra\fcdncw\flcd 19]408029.doc 16pixel backward and down is a foreground one IF (PreviousPixel.TopRightQuadrant GREATER THAN

THEN

pixel backward and up is a foreground one CurrentPixel.BottomLeftQuadrant 0

ENDIF

ENDIF

ENDIF

ELSE

Gap is horizontal Similar code to the vertical case is inserted here

ENDIF

ENDIF

T Junctions Referring to Figs. 16A and 16B, there is illustrated a junction between two or more pixel lines intersecting or meeting at right angles. This like" junction occurs in various characters including characters such as and but is not limited 20 thereto. In T-Like junctions 1700, the uncorrected output 1701 results in one or more quadrants 1702 undesirably switched to foreground colour and if the uncorrected output were to be displayed, without correction, the corresponding output text would appear lumpy or blobby.

The feature extraction unit 504 detects T-like junctions from the modified data representation and passes this information as well as a corresponding uncorrected output for the T-like junction, so detected, to the data generation unit 505. The data generation unit 505 amends (corrects) the uncorrected output to a desired corrected output. For example, Fig. 16B illustrates two case examples of T-like junctions. In the example of Case 1, a T-like junction 1703 input to the smooth processing system 500 is 30 modified by the data modification unit 502 and stored as a modified data representation on a data line store 503. The feature extraction unit 504 detects (identifies) the input Tlike junction 1703 from the modified data representation and outputs to the data generation unit 506 an uncorrected output substantially similar to the uncorrected output 1701 of Fig. 16A. The data generation unit 505 removes (corrects) the one or more undesired quadrants 1702, and outputs a corrected output 1704 for the corresponding input T-like junction 1703. The example in Case 2 shown in of Fig. 16B can be described substantially as Case 1 example, and variations on these examples can be made without departing from the spirit or scope of the invention.

FLCD19 408029 CFP0757AU [o:\cisra\flcdncw\tlcd l9408029.doc -17- A set of rules for detecting (identifying) a T-like junction will now be described with reference to Fig. 17 and Fig. i8. In Fig. 17 there is illustrated two examples of T-like junctions expressed in a modified data representation excepting that where a quadrant value is 1.0 (one) or greater, the quadrant is indicated by the colour black to provided a visual representation of an input character represented by the modified data.

The set of rules for identifying a T-like junction are set out as follows and described with reference to a first example of Fig. 17: a current pixel 1800 in the modified data representations has a total quadrant value of 1.0 (one); (ii) the current pixel 1800 also consists of at least one quadrant 1801 having a value of 0.5 (half); (iii) at least one pixel 1802, on a current scan line adjoining the current pixel 1800, has a total quadrant value of 4.0 (four) or greater; and (iv) a "next" pixel 1803, immediately adjacent to the at least one pixel 1802, comprises quadrant values substantially similar to the current pixel 1800 excepting that the quadrants of the next pixel 1803 are arranged as a permutation of the quadrant of the current pixel 1800.

Typically, the permutation of the quadrants referred to in rule (iv) above includes of a reflection of the quadrants about a centre axis of the pixel. For example, the 20 second pixel 1803 is a reflection of the current pixel 1800 about a centre vertical axis.

A second example illustrated in Fig. 17 can be described substantially as that of the first example, excepting that one pixel 1804 is a reflection of another pixel 1805 about a centre diagonal axis.

Preferably, the above set of rules for identifying a T-like junction is applied by the feature extraction unit 504 to pixels on a current scan line of the modified data representation. A T-like junction is identified, in accordance with the present embodiment, when each one of the previously described rules are satisfied.

Turning now to Fig. 18, there is shown an example of a pixel arrangement where SOone of the four rules for identifying a T-like junction is not satisfied. One pixel 1900 is 30 not a permutation of a quadrant of another pixel 1901 in accordance with that described by rule Typically, the set of rules for identifying a T-like junction is applied to pixels belonging to a horizontal scan line of pixels and therefore "horizontal" T-like :junctions are identified. Horizontal T-like junctions include characters such as and "+"(plus sign) where an application of the set of rules is performed to pixels adjacent in a horizontal direction. Alternately, vertical T-like junctions such as those encountered in the letter can be identified by a application of the set of rules in a vertical scan line. However, in the present embodiment vertical T-like junctions are dealt with as a plurality of right angle joints described hereinafter.

FLCD19 408029 CFP0757AU [o:\cisra\flcdncw\kd 19]408029.doc -18- Although detecting and smoothing horizontal T-like junctions improves the desired text quality sharpness), character such as the letter are noticeably less smooth than often desired.

Right Angle Joints Yet another feature that is identified by the feature extraction unit 504 is its ability to account for a Right-Angle Joint. Fig. 19A illustrates an example of a Right-Angled Joint 2000 encountered in text characters such as A rotation or reflection of the Right-Angled Joint 2000 is also encountered in a multitude of letters including and but not limited thereto.

The feature extraction unit 504 identifies a Right-Angled Joint when the following rules are satisfied for a one or more pixels: 1) a current pixel has a total quadrant value of 1.0 (one); and 2) a single quadrant of the current pixel has a value of 0.5. If the above two rules for detecting (identifying) a Right-Angled Joint are satisfied a further check is applied to the current pixel: 3) if the single quadrant, of the current pixel, having a value of 0.5 is a left quadrant, determine (or check) a total quadrant value of an adjacent pixel to the left of the current pixel; or 4) if the single quadrant, of the current pixel, having a value of 0.5 is a right 20 quadrant determine (or check) a total quadrant value of an adjacent pixel to the right of the current pixel.

In either rule 3) or rule 4) above, if the determined total quadrant value of the adjacent pixel is 5.0 (five) then the feature extraction unit 504 signals the data generation unit 505 that a Right-Angle Joint has been detected (identified) and the data generation unit 505 sets all the quadrants of the current pixel to 0 (zero).

S. The above described rules for detecting Right-Angled Joints can identify a plurality of rotations and reflections, of which the Right-Angled Joint 2000 of Fig. 19A is an example. Also shown in Fig. 19A is an uncorrected output 2001 corresponding to the Right-Angle Joint 2000 input to the smoothing processing system 500.

Referring now to Fig. 19B there is illustrated two case examples of Right-Angle Joints and their corresponding corrected output. In Case 1 a Right-Angle Joint 2002 is input to the smoothing processing system 500, a corresponding uncorrected output substantially similar to the uncorrected output 2001 of Fig.. 19A is output by the feature extraction unit 504 to the data generation unit 505 where the uncorrected output is amended (corrected) to a corresponding corrected output 2003. The corrected output 2003 is then directed to a display device. Case 2 example relates to an example substantially similar to Case 1.

The identifying and correcting of features such as Right-Angled Joints also smooths characters having "vertical" T-like junctions described previously.

FLCD19 408029 CFP0757AU [o:\cisra\ficdnc\flcd 19408029.doc -19- Angled Lines Several text character font types (including Chinese, Japanese and Greek character fonts) make extensive use of angled lines, many of which are two-pixels wide. Generating a modified data representation for a two-pixel wide angled line and a corresponding uncorrected output from the feature extraction unit 504 has the effect of thickening the line. Often, thickening the lines in text characters (letters) is not desired nor was it intended by the font designer.

Identifying angled lines (typically two pixel wide) such as those found in characters like (an in script font) and other characters allow the smoothing processing system 500 to correct unnecessarily thickened angled lines.

When an angled line is detected, by the feature extraction unit 504 from the modified data representation, the data generation unit 505 is signalled to thin the angled line according to a smoothing requirement (desired visual appearance). To detect angled lines the feature extraction unit 504 identifies pixels having a total quadrant value of 4.5 (four and one half). Having identified a current pixel with a total quadrant value of 4.5, pixels neighbouring the current pixel are checked to ensure that the current pixel is: 1) not at a top or bottom of a vertical line, by checking that at least one adjoining pixel on either the left or right side of the current pixel is of 20 foreground colour; and not at the point of intersection of a T-like junction, by checking that only one adjacent pixel, either to the left or to the right, is of foreground colour.

~The above two conditions are checked because a pixel in a modified data representation can have a value of 4.5 at the beginning and end of a vertical line (or stroke) and at a T-like junction.

Having satisfied the above two conditions, for ensuring that a pixel having a total S quadrant value of 4.5 is part of an angled line, the feature extraction unit 504 signals the data generation unit 505 which, in turn, changes a quadrant of the current pixel according to the following rules: 30 If a foreground colour quadrant of the current pixel abuts at least two background coloured quadrants of adjacent pixels, the foreground colour quadrant is turned to a background colour, and the foreground colour quadrant that is to be turned to the background colour must not adjoin a quadrant of an adjacent background pixel that is to be turned to a foreground colour.

Fig. 20A illustrates an example of a two-pixel wide angled line 2100 input to the smooth processing unit 500 and a corresponding uncorrected output 2101. Fig. shows again the two-pixel wide angled line 2100 of Fig. 20A and the corresponding corrected output 2102. A required amendment from the corresponding uncorrected FLCDI9 408029 CFP0757AU [o:\cisa\flcdnew\fldl9]408O29.doc output 2101 of Fig. 20A to the corresponding corrected output 2102 of Fig. 20B is performed in the present embodiment by the data generation unit 505.

The features previously described under the heading "Features To Extract" including: Single Pixel Hole; Pixel Wide Gaps; T-Like Junctions; Right Angle Joints; Angled Lines in no way represent an exhaustive list of possible features that can be used in conjunction the smoothing processing system 500 (or detected by the feature extraction unit 504).

Smooth Processing System of a Second Embodiment A second embodiment of the present invention is illustrated in Fig. 21 and shows a smooth processing system 2202 which includes substantially as described in the first embodiment, excepting that a pixel line store 2200 and a data combining unit 2201 have been further added to the apparatus of the first embodiment (Fig. The second embodiment, with the additional pixel line store 2200 and data combining unit 2201, allows text characters on a no-flat background to be smoothed effectively without substantial undesirable disruption to the background effect. Examples of non-flat backgrounds include: natural images as background with text characters to be smoothed overlayed on the natural images; or an embossed effect on a background image with characters to be smoothed displayed over the embossed effect.

Preferably, a zone segmentation unit (as described above) identifies data S 20 comprising text characters on a non-flat background with the data so identified being input to the smooth processing system 2202 of the present embodiment.

Input data, comprising text data characters to be smoothed, is received by the data modification unit 502 and the pixel line store 2201 via an input 501. The input data to U o the data modification unit 502 follows substantially the same process as that described for the first embodiment, however a corrected output from the data generation unit 505 Sois now input to the data combining unit 2201. Synchronised with the input of corrected output data, received from the data generation unit 505, the data combining unit 2201 receives (reads) from the pixel line store 2200 corresponding input data on a pixel by pixel bases.

30 The data combining unit 2201, by comparison of the two input data or otherwise, determines whether a current pixel is to remain substantially the same colour as it had at input, to the smooth processing unit 2202, or a quadrant of the current pixel is to change colour. In the event that the current pixel is to remain the same colour as it had at input to the processing unit 2202, a corresponding pixel of the pixel line store 2200 is output to a display device. In an event that a foreground coloured quadrant is to be changed to background colour, neighbouring quadrants having background colour values from the pixel line store 2200 are used to determine a colour of the quadrant to be so changed. Otherwise the corresponding pixel in the pixel line store 2200 is modified on output in accordance with that dictated by the data generation unit 505 so FLCD19 408029 CFP0757AU [o:\cisr;\fcdiew\lcd9]4O8029.doc 21 that the required one or more quadrants of the current pixel are set to a foreground colour.

Smooth Processing System of a Third Embodiment A third embodiment of the present invention can be described substantially as for the previous two embodiments, excepting that the modified data representation need not be described by floating point numbers (eg. 1.0, 1.25). A modified data representation including five bits, that can be set either on or off one can be used for each pixel. For example, Fig. 22 illustrates a pixel 2300 having a foreground colour quadrant 2301 and three background coloured quadrants 2302 which can be represented by a modified data representation having of five bits 01010. A first bit (from left to right) of the present example is a zero, indicating that not all quadrants of the pixel 2300 are foreground colour. Alternately, if the first bit is set on then all the quadrants of the pixel 2300 are foreground colour. A second bit of the-example is one (on) indicating that a first column 2303 of a two by two array of quadrants for the pixel 2300 contains a quadrant that is of foreground colour. A next (third) bit of this example is zero indicating that a second column 2304 contains no quadrant of foreground colour. A fourth bit of the example is one (on) indicating that a first row 2305 of the two by two array of quadrants for the pixel 2300 contains a quadrant that is of foreground colour and fifth bit is zero indicating that a second row 2306 contains no 20 quadrant of foreground colour. The combination of five bits as 01010, of the present example, indicate a quadrant 2301 at the intersection of the first column with the first *row is of foreground colour.

A combination of the first bit in the five bit representation with the remaining four bits allows the five bit representation to efficiently include additional information about neighbouring pixels. In particular, a. five bit representation of a current input pixel includes information as to whether adjoining quadrants of neighbouring pixels are of foreground or background colour. Consequently, a five bit representation has the advantage of minimising the storage requirements for the modified data representation.

:Changes to the modified data representation can be made without departing from the S 30 scope and spirit of the invention. For example, integer values rather than a five bit representation (or floating point number representation) can be used to identify which quadrant is foreground colour in a current pixel.

Although each of the embodiments have been described herein with reference to quadrants, where each pixel typically comprises four quadrants, and thus a higher resolution of double the normal resolution twice the number of independently controllable picture elements per inch per dimension), the process of each embodiment can be implemented with other than double the resolution without departing the scope or spirit of the invention.

FLCDI9 408029 CFP0757AU [o:\cisra\flcdnew\flcd 19]408029.doc -22- Fig. 23 illustrates an example of a three times increase in resolution, from a normal resolution to a higher resolution. There is shown a first image 2400 comprising four pixels at normal resolution, a second corresponding image 2401 at double the normal resolution and smoothed in accordance with any one of the previous embodiments.

A third image 2402, corresponding to the first image 2400, at three times the normal resolution and smoothed in accordance with any one of the previous embodiments is also shown in Fig. 23.

In each of the first, second and third images of Fig. 23, diagonal dotted lines indicate a desired image intended for display but not, exactly, achievable because of the discrete nature of a pixel based image. A smooth processing system according to any one of the previous embodiments approximates the desire image at a predetermined resolution. At double the normal resolution the desired image of the present example (Fig. 23) is sufficiently approximated by switching to foreground colour two pixels (quadrants) 2402, each quadrant equivalent to one quarter of a corresponding pixel at the normal resolution. At three times the resolution the desired image can be better represented by switching to foreground colour a plurality of pixels 2405.

Consequently, while a pixel at a normal resolution is represented by four quadrants in a modified data representation, a predetermined number of pixels are generated (or S. 20 corrected) by a data generation unit 505 consistent with a desired output.

*Thus in Fig. 23, produced by a data generation unit 505, the desired output for a quadrant marked as foreground colour by a data modification unit in cooperation with a feature extraction unit; is illustrated by the second corresponding image 2401. At three times the resolution the third image 2402 adequately approximates the desired output.

The described embodiments can be implemented in software, for example on a IBM PC/AT type computer, or alternately in hardware, for example as part of a specific display device. Further implementations can include hybrid software and hardware arrangements. In software implementations, computer code can be provided S° by means of storage devices such as magnetic and optical disks, or via computer networks, such as the Internet for example.

The foregoing describes only a number of embodiments of the present invention, however, modifications and/or changes can be made by those skilled in the art without departing from the scope and spirit of the invention.

.9o FLCDI9 408029 CFP0757AU [o:\cisra\lcdnw\jcd 19408029.do

Claims (22)

1. A method of extracting information from a bitmap image representation, the bitmap image representation comprising at least a portion of a bitmapped character formed from a plurality of pixels, said information being inherent in the arrangement of pixels of the bitmapped character, said method comprising the steps of: generating a modified data representation for the bitmap image representation wherein said modified data representation is characterised by additional information being inherent in the arrangement of pixels in the bitmap character; and identifying one or more predetermined features from said modified data representation.

2. The method according to claim 1, wherein the modified data representation enhances the bitmap representation to facilitate the extraction of the implied information.

3. The method according to claim 1, wherein said method comprises the further consecutive steps of: amending said modified data representation in accordance with a predetermined set of rules for each feature identified; and creating an altered bitmap image representation based on the amended modified data representation.

4. The method according to claim 3, wherein the bitmap image representation is provided at a first image resolution and the altered bitmap image representation is created 25 for a second (higher) resolution. o

5. The method according to claim 1, wherein said method further comprises the steps of: 3creating an uncorrected bitmap image representation based on the modified data representation; correcting said uncorrected bitmap image representation in accordance with a "predetermined set of rules for each feature identified; and providing for display on a display device the corrected bitmap image representation.

6. A method of enhancing a bitmapped character, said bitmapped character comprising a plurality of pixels, said method comprising the steps of: 408029 CFP0757AU [o:\cisra\flcdnew'flcd19]408029.doc -24- inputting a bitmap image representation including at least a portion of the bitmapped character; generating a modified data representation for each pixel of the bitmap image representation wherein said modified data representation is characterised by additional information being inherent in the arrangement of pixels in the bitmap character and storing said modified data representation; identifying one or more predetermined features relating to said bitmapped character from the stored modified data; and generating an output bitmap image representation based on the modified data representation and the identified predetermined features.

7. A bitmapped character enhancing apparatus comprising; input means for inputting a bitmap image representation comprising at least part of a bitmapped character; data modification means connected to said input means for providing a modified data representation wherein said modified data representation is characterised by additional information being inherent in the arrangement of pixels in the bitmap character; data storage means, coupled said data modification means, for storing said modified data representation; feature extraction means for extracting from the modified data representation predetermined features of said bitmapped character; and data generation means for providing an output bitmap image representation based on the modified data representation and the predetermined features; S 25

8. A bitmapped character enhancing apparatus comprising; 9• input means for inputting a bitmap image representation comprising at least part of a bitmapped character; data modification means connected to said input means for providing a modified data representation of said bitmapped character wherein said modified data representation 30 is characterised by additional information being inherent in the arrangement of pixels in the bitmap character; o ~data storage means, coupled said data modification means, for storing said modified °data representation; feature extraction means for extracting from the modified data representation of predetermined features of said bitmapped character; S-data generation means for providing a corrected bitmap image representation of said S.bitmapped character based on the modified data representation and the extracted Spredetermined features; [o:\cisra\flcdnew\flcdI 9]408029.doc pixel data storage means, coupled to the input means for receiving and storing said input bitmap image representation; and data combining means for receiving the input bitmap image representation and corrected bitmap image representation, from the pixel data storage means and said data generation means respectively, and for producing an output bitmap image representation based on the input bitmap image representation and corrected bitmap image representation.

9. The invention according to claim 6, 7 or 8, wherein said input bitmap image representation is intended for display at a first resolution and said output bitmap image representation intended for display at a second (higher) resolution.

The invention according to any one of the preceding claims, wherein said bitmapped character includes text characters and/or line drawings.

11. The invention according to claim 10, wherein the text characters include oriental and occidental text characters.

12. The invention according to any one of the preceding claims, wherein said modified data representation comprises a five bit representation for each pixel of the bitmap image.

13. The invention according to any one of the preceding claims, wherein said modified data representation comprises floating point numbers representing each pixel of the bitmap image. S

14. The invention according to any one of claims 1 to 12, wherein said modified data representation comprises integer numbers representing each pixel of the bitmap image.

15. The invention according to any one of the preceding claims, wherein said e 30 predetermined features are selected from the group consisting of a pixel hole feature, a T- like junction feature, a right angle joint feature, and an angled line feature. S..

16. A method of enhancing a bitmapped character, said method substantially as oo" described herein with reference to the accompanying drawings.

17. The invention as claimed in any one of claims 12, 13 or 14, wherein said modified •r _data representation for a particular input pixel comprises values for a plurality of output 19 408029 CFP0757AU [o:\cisra\flcdnew\flcd 19]408029.doc -26- pixels collectively corresponding to said input pixel and values for adjacent one or more input pixels of correspondingly adjacent pixels to said input pixel.

18. A method of displaying a bitmapped character formed from a bitmapped image representation including a plurality of pixels intended for display at a first display resolution, said method comprising the steps of: generating a modified data representation for the bitmapped image representation, said modified data representation comprising a plurality of pixels at a second display resolution for each pixel at the first display resolution wherein said modified data representation is characterised by additional information being inherent in the arrangement of pixels in the bitmap character; and identifying from said modified data representation at least one predetermined feature of said bitmapped character being inherent in the arrangement of pixels; altering said modified data representation in accordance with a predetermined feature; creating an altered bitmap image representation of said bitmapped character based on the altered modified data representation; and displaying the altered bitmap image representation with a display device at the second display resolution.

19. A method according to claim 18, wherein step comprises the sub-steps of: (ba) allocating values to said pixels at said second display resolution, said values being determinable based upon: a positional relationship between said pixels at said second display 25 resolution and a particular pixel at said first display resolution, (ii) a difference in pixel colour between said particular pixel and adjacent .5 pixels at said first display resolution; S(bb) comparing said values of adjacent pixels at said second display resolution in accordance with at least one predetermined set of identifying rules: and (bc) based on the comparison, assigning a particular predetermined feature to said .fee adjacent pixels. 5.55 ••go

20. A method according to claim 19, wherein for a particular pixel at said first display resolution, said values are allocated to the corresponding pixels at said second display resolution, and to pixels immediately adjacent to said corresponding pixels at said second S° display resolution. °oooo 408029 CFP0757AU [o:\cisra\flcdnew\flcd 19]408029.doc -27-

21. A computer readable medium comprising computer implementable instructions that carry out the method of any one of claims 1 to 6 and 18 to

22. A computer apparatus configured to implement the method of any one of claims 1 to 6and18 to DATED this twenty-eighth Day of March 2000 Canon Kabushiki Kaisha Patent Attorneys for the Applicant SPRUSON FERGUSON 9. 99 9 9 9 9 9 9 408029 CFP0757AU 40802 CFFO57AU[o:\cisra\flcdniew\flcdl 9]408029.doc