JPH09330418A - Binarizing converter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the omission of information caused by binarizing conversion by composing third edge data with respect to one value of binarizing information adjacent to third edge data on temporarily binarized picture information by the other value. SOLUTION: Processing objective picture data for one picture is fetched into a video RAM 408 to execute luminance data conversion. Then a CPU 401 inspects the luminance distribution for one picture from this luminance data to detect maximum luminance level data being a luminance level showing the distribution of the maximum luminance distribution. The detected luminance information is converted to one value of binarized information and the other detected luminance information is converted to obtain temporarily binarized picture information. With respect to one value of the binarized information adjacent to third edge data not including binarized information on the temporarily binarized picture information, third edge data are composed by the other value on the tenprarily binarized picture information. Thereby the omission of information caused by binarizing conversion is reduced.

Description

Detailed Description of the Invention

[0001]

[0001]

[0002]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a binarization conversion device which performs binarization conversion from a multicolor image having density information.

[0003]

[0002]

[0004]

2. Description of the Related Art When an image is displayed on a black and white display unit, there is a gradation display in which the density of black and white data is changed according to the brightness level of an original color image and the image is displayed by shading.

[0005]

Conventionally, various image processing methods for color still images have been proposed. However, based on the information of color image data or black-and-white multi-gradation data, black-and-white prints and images on a monochrome display. Is obtained, conversion processing is performed by various methods including binarization processing.

[0006]

Therefore, in general, a certain threshold value is set, the binarization process is performed with the level as a reference, and the contour (edge) is extracted.

Conventionally, when a document image or the like is transmitted by a device such as a facsimile, the document image or the like is divided into a plurality of pixels, and the density level of each pixel is converted by an optical means into an analog signal corresponding to each density. Further, the image information is processed by converting the electric signal into a binary digital signal using a threshold value.

[0008]

FIG. 7 is a block diagram showing a binarization process of a conventional image reading apparatus.

In the figure, image data of characters and figures are
An input signal 501 as image data from a video output device (not shown) such as a video tape recorder or a personal computer.
Next, it is input to the memory 502. In the memory 502, it is stored as a quantized value using, for example, an 8-bit binary number. Here, the information of the peripheral pixel density level of the pixel to be binarized is sent to the average calculation unit 503, and the average value thereof is calculated.

[0010]

In the threshold value determining unit 504, the average calculating unit 50
The threshold value used for binarization corresponding to the average value signal calculated in 3 is output to the comparator 506. Here, in the threshold value determining unit 504, the average value calculated by comparing the average value signal calculated by the average calculating unit 503 with the boundary value signal 505 set based on the pixel density level in a predetermined range is the boundary value. When the calculated average value is smaller than the boundary value, the upper correction value is set below the pre-standardized threshold value. A threshold is used. In addition, the comparator 506
Then, the quantized value stored in the memory 502 is compared with the input threshold value, converted into a binarized signal, and output. In this way, the threshold value determining unit 504 obtains a more appropriate binarized output by appropriately correcting the threshold value supplied to the comparator 506 according to the surrounding pixel density level.

[0011]

[0007]

[0012]

In the conventional method, when there are adjacent data whose brightness levels are close to each other,
In color image data or multi-gradation data, it was possible to distinguish by color or gradation, but in the case of a single gradation of black and white display, there is a problem that it cannot be clearly read.

Further, on such a brightness level,
A method may be considered in which another threshold value is provided at a position where adjacent data can be separated and the data sandwiching the threshold value is inverted and converted. Therefore, an object of the present invention is to provide a binarization conversion device that reduces information loss due to the binarization conversion.

[0014]

[0008]

[0015]

A binarization conversion device according to claim 1 is a binarization conversion device for converting luminance image information having at least luminance information into binary image information.
The brightness distribution detecting means for detecting the brightness distribution of the brightness image information, the maximum brightness level information detecting means for detecting the most brightness level data which is the brightness level showing the distribution with the most brightness distribution, and the most brightness level information are binarized. A conversion unit for converting the brightness information having the other brightness into one value of the information into the other value of the binarized information to obtain the temporary binarized image information; and edge data in the brightness of the brightness image information. First edge data generating means for generating one edge data;
Second edge data generating means for generating second edge data which is edge data in the binarized information of the binarized image information, and third edge data which is present in the first edge data and is not present in the second edge data. The third edge data generating means for generating the edge data and one value of the binarized information adjacent to the third edge data on the temporary binarized image information are used to temporarily store the third edge data in the other value. And a synthesizing means for synthesizing the binarized image information.

A binary conversion device according to a second aspect of the present invention is the binary conversion device according to the first aspect, wherein the third edge data generating means excludes the first edge data and the second edge data. The third edge data is generated by taking a logical OR, and the synthesizing means synthesizes the temporary binary image information and the third edge data by taking an exclusive OR.

[0017]

[0009]

[0018]

In the binarization conversion device according to the present invention, the maximum luminance level information is converted into one value of the binarized information, and the luminance information having the other brightness is converted into the other value of the binarized information, and the temporary 2 It is possible to obtain the binarized image information, prevent the information from being lost by a plurality of binarization processes, and invert and display the binarized image having a large area.

[0019]

[0010]

[0020]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described with reference to FIGS.
This will be described with reference to FIG. 1 and 2 are examples showing a method of an embodiment of the present invention. FIG. 3 is a block diagram showing the configuration of a control device that implements and operates software for performing the processing of the present invention. In FIG. 3, 401
Is a CPU (Central Processing U)
ROM (Read) via the bus 413.
Only Memory 402, RAM (Rand)
om Access Memory (403) to control the peripheral devices 404, 408, 409, 411. The computer program of the software is stored in the ROM 402.

[0021]

The image data to be processed is obtained from a CD-ROM, a floppy disk or a communication line (not shown), and the processed image data is stored in the portable terminal I / F section 41.
1 through PDA (Personal Digital
l Assistant) 412 and other portable terminals. The captured image data is captured in the video RAM 408. The main operation is performed from the pointing device 410 such as a mouse through the pointing device control unit 409.

[0022]

When the image signal of FIG. 1A is a color signal, the CPU 401 re-stores in the video RAM 408 the brightness conversion image converted from the color signal to the brightness signal by the software of the ROM 402.

Next, the frequency of appearance of the brightness data with respect to the brightness is measured by software, and the brightness distribution as shown in FIG. 1B is examined. The analysis result is stored in the RAM 403 as peak brightness.

[0024]

Next, the maximum luminance among the luminances stored in the RAM 403 is selected as the reference value ref, and the difference between the luminance of the reference value and the adjacent peak luminance and the luminance of the reference value is selected. Is calculated by software, RAM4
03. This luminance difference is 2 as shown in FIG.
Used as the range rng for quantification.

[0025]

Next, the brightness data of FIG. 1A is binarized and converted by software using the above-mentioned reference value ref and range rng to obtain FIG. That is, the brightness is re
The data within f ± rng / 2 is stored in another frame of the video RAM 408 as 0 (black) and the other data as 1 (white). In the example of FIG. 1C, as can be seen from FIG. 1D, only the blue color of the sea is 0 (black) and the other portions are 1 (white).

[0026]

Next, the image of FIG. 1E is obtained from the binarized image data by edge detection (outline extraction). That is, the boundary between the sea and other parts is obtained as a contour. Edge detection is performed as follows, for example. The edge refers to a boundary where areas of different colors are adjacent to each other. Therefore, if the difference between the image luminance data in one horizontal scanning period and the data obtained by delaying the data by one pixel or several pixels is obtained, the luminance change data in the horizontal scanning period can be obtained.

[0027]

For the vertical scanning period, the brightness change data in the vertical direction can be obtained in the same manner by using a signal delayed by one horizontal scanning period or several horizontal scanning periods. Here, this processing is also performed by software. Ie video RAM
Software processing is performed from the binarized image data accumulated in 408, and stored in another frame of the video RAM 408.

[0028]

A computer program is stored in the ROM 402, and the CPU 401 executes the software processing by using the RAM 403 and the video RAM 408 according to the program of the ROM 402 via the bus 413.

[0029]

Then, the brightness-converted image to be processed is edge-detected in the same manner to obtain the image data of FIG. 1 (f). This is also done by software processing, and the video RAM 40
8 in yet another frame. This Figure 1 (f)
Means that the contour of FIG. 1A has been extracted.

[0030]

Next, FIG. 1 stored in the video RAM 408.
If the exclusive OR of the image data of (e) and the image data of FIG. 1 (f) is performed by the software processing, as shown in FIG.
Data existing only in one of (e) and FIG. 1 (f) can be extracted.

[0031]

Finally, the exclusive OR of the binarized image data of FIG. 1 (c) and the edge data of FIG. 2 (g) is software processed to obtain the image data of FIG. 2 (h).
That is, in the image of FIG. 2H, the road portion is displayed in black on the land side and white on the sea side in the binarized image of black of the sea and white of the land, and necessary information is obtained as a binarized image. be able to.

[0032]

In this way, the CPU 401 controls the ROM 4
02 software, the image data accumulated in the video RAM 408 is binarized through the RAM 403, the result is accumulated in the RAM 403, and the TV signal is transmitted through the video control circuit 404 according to the instruction of the pointing device 410. Output to encoder 405,
The video output DAC 406 outputs the analog signal to the monitor 407.

[0033]

Here, FIG. 1B shows the information of FIG. 1A with the horizontal axis representing the luminance level and the vertical axis representing the appearance frequency. For example, maps are generally represented by outlines of symbols that represent roads, railways, stations, contour lines, sea lakes, rivers, etc., characters that show proper nouns of each, and color divisions that represent cities, towns, wards, towns, and their boundaries. ing. In the case of FIG. 1 (a),
Shown is the case where the sea is shown in blue, the road is shown in gray, the land is shown in light brown, and it is displayed in black letters.

[0034]

Characters, boundaries, railroads, main roads, etc. are often expressed in dark black, red, blue, etc., general roads are gray, sea lakes and rivers are light blue, park green areas are light green, etc. expressed.
In addition, buildings and the like surrounded by contour lines are often painted in different colors to distinguish them from each other.

Therefore, the brightness-frequency distribution of the map is, for example,
An example as shown in FIG. 1B is obtained. In addition, when the input data is obtained from a commercial map or the like with a scanner or the like, for example, there may be a case where the originally existing information such as dust or added data is not included.

[0036]

However, roughly speaking, if the above description is made, the part having the highest brightness corresponds to white, the part having the lowest brightness corresponds to black, and the black part has a contour line or a boundary line and character information. The gray area is the general road, and the white area is the distribution map of the color area representing the area division such as other rivers, lakes and rivers.

[0037]

[0025] Here, contour lines are used for most boundaries in normal maps, but examples are considered such as sea-land boundaries where contour lines are omitted in peripheral maps of road maps and tourist maps. To be In such a case, for example, if a blue portion of the sea or the like and a light brown portion of the land are discriminated and binarized, portions lower than the blue and blue luminance levels are represented by black, and portions other than that are represented by white. . Then, in the sea filled with blue, all the parts such as bridges and letters in which the brightness is lower are black, and the information is lost.

[0038]

In such a case, the luminance as a color-
In the frequency distribution, frequencies of blue, light brown, gray, and black appear in order of frequency. The order of brightness levels is light brown, blue, gray, and black. Therefore, FIG. 1C shows that the brightness level of blue is binarized to black and the others are binarized to white. This is the temporary binarized image information, and the luminance-frequency distribution before the binarization is shown in FIG.

As shown in FIG. 1D, the range of binarization to the black level is the range of the difference between the brightness of blue and the brightness of gray adjacent to the center of blue. Therefore, in this case, the range of the difference in brightness between blue and grey, centered on blue, is black, and only blue is replaced with black and the others are replaced with white.

[0040]

FIG. 1E shows the result of edge detection from FIG. 1C, and FIG. 1E shows the result of edge detection from FIG. 1A of the display example to be converted. f).

Then, the roads and characters missing in FIG. 1 (e) appear in FIG. 1 (f). Therefore, in FIG.
If the exclusive OR of (e) and FIG.
Information in either (e) or FIG. 1 (f) is obtained as FIG. 2 (g). Then, if the exclusive OR of FIG. 2 (g) and FIG. 1 (c) of the previous temporary binarized image information is taken, roads and characters in the land portion are added to FIG. 1 (c). FIG. 2 (h) is obtained.

[0042]

FIG. 6 shows an example of the above-mentioned luminance-frequency distribution. FIG. 6A shows the case of a yellow monochrome screen as shown in FIG.
FIG. 6B is a diagram schematically showing a case of black characters or a black frame on a white background, and FIG. 6C is a diagram schematically showing a case of multicolor complicated display. Each,
The frequency peak numbers are 1, 2, and 10.

[0043]

Next, an example of a flowchart showing the binarization processing method of the present invention is shown in FIGS. STEP1
Brightness distribution detection in STEP 5, STEP 6 to STEP 13
In step 14, the second edge data is generated in STEP 14, the first edge data is generated in STEP 15, and the third edge data is generated in STEP 16.
At 17, the third edge data and the temporary binarized image information are combined.

[0044]

From STEP 1 start, first, STEP 2
Then, one screen of image data to be processed, for example, the data shown in FIG. 1A is loaded into the video RAM 408, and brightness data conversion is performed in STEP3. The luminance data conversion is performed, for example, when the image data to be processed is RGB data, Y = 0.2988 × R + 0.5868 × G +
The luminance component Y is obtained from 0.5868B, and in the case of YUV data, Y
Use only data.

[0045]

The brightness data thus obtained is A. Next, in STEP 4, the luminance distribution analysis is performed from the luminance data for one screen. That is, the luminance is plotted on the horizontal axis and the frequency of appearance on the vertical axis, and the luminance distribution for one screen is examined. This STEP corresponds to FIG. In STEP 5, Y ₁ (Pmax) having the highest frequency is set as the reference value ref based on the result of the luminance distribution analysis (FIG. 1D).

[0046]

In STEP 6, the number of local peak values of frequency is checked from the result of the luminance distribution analysis, and it is checked whether the number is 2 or more. If Yes, the process proceeds to the setting of the range rng in STEP 7, and the difference between the highest frequency luminance Y ₁ and the frequency peak luminance Y ₂ closest to Y ₁ on the luminance axis is set as the range rng. For example, in the case of FIG. 1 (d),
Since Y ₁ corresponds to the brightness of blue in the sea and the brightness of ash corresponds to Y ₂ , the difference between the brightness levels of blue and ash is set as the range rng. On the other hand, when the result in STEP 6 is No, the frequency peak is 1, which corresponds to the case where the entire screen is composed of a single color.

In this case, in STEP 8, the difference between the highest luminance and the lowest luminance is set as the range rng.

[0048]

Next, in STEP 9, the brightness of the reference value Y ₁
Is the central brightness, that is, the difference between the maximum brightness and the minimum brightness is 1.
Check if it is higher than / 2, and if Yes,
The parameter P used later is set to 0 (STEP1
0). If No, P is set to 1 in STEP 11.

[0049]

Next, in STEP 12, each data of the brightness data A is changed to Y ₁ ± by using the reference value ref, the range rng, and the parameter P set in the above.
If it is within rng / 2, it is set to P, and the other data is set to the inverted value of P.

The image data thus obtained will be referred to as binarized data B. This binarized data B becomes temporary binarized image information.

[0051]

At STEP 13, it is checked whether the number of peaks having a frequency equal to or higher than the specified value is 2 or more. The specified value is set, for example, 70% or more of the frequency of the maximum brightness, or the like, from the visual effect. If Yes, in STEP 14, edge detection is performed from the binarized data B, and the binarized data C is detected.
And This binarized data C becomes the second edge data. In FIG. 1, it corresponds to (e).

[0052]

In STEP 15, edge detection is performed from the brightness data A to obtain binarized data D. This is shown in FIG.
Corresponding to (f), the binarized data D becomes the first edge data.

Further, in STEP 16, the binarized data C
And the binary data D are subjected to exclusive OR (EXOR) processing, and the obtained data is used as the binary data E. The binarized data E becomes the third edge data and corresponds to FIG. Further, in STEP 17, the exclusive OR (EXOR) processing of the binarized data B and the binarized data E is performed, and the obtained data is set as the binarized data F. STEP17 is
The synthesis of the third edge data and the temporary binarized image information is performed, and the resulting binarized data F corresponds to FIG.

[0054]

As a result, the binarized data is obtained by exclusive OR processing of the edge data of the binarized data C that has undergone the luminance range processing and the binarized data D of the edge data obtained from the original luminance data. Edge data that does not appear in D is obtained as binarized data E, and the desired binarized data F is obtained by exclusive OR processing of the data with the binarized data B that is the binarized data of the area. Will be obtained.

[0055]

STEP 18 shows the case of No in STEP 13, which is the case where the number of peaks having the brightness equal to or higher than the specified value is 1 or 0. In this case, the binarized data B is output.

In STEP 19 and thereafter, whether the output is an inverted output or not is selected, and in STEP 19 it is selected whether to output the output as it is or the output is made as an inverted output according to the setting or the instruction.
After P21, the flow ends and STEP22 ends.

[0057]

As described above, the processing has been described as the processing for one screen, but the masking processing is performed in a part of the process of the processing, that is, the above processing is partially performed by designating a partial area of the image and fitted. It is also possible to obtain a desired image by performing. Further, even if the image to be processed is a binary image or another format such as vector processing data, it is converted into a format usable for the main conversion processing by the pre-processing and converted into another format by the post-processing. By doing so, it can be applied to image data of other formats.

[0058]

[0040]

[0059]

As described above, in the binarization conversion device of the present invention, the loss of information due to the binarization conversion can be reduced, and the binary data area having a large area can be displayed inverted. After conversion, it is possible to obtain a more effective display visually.

[Brief description of drawings]

FIG. 1 is a part of a schematic diagram showing a processing situation of an embodiment of the present invention.

FIG. 2 is a part of a schematic diagram showing a processing situation of an embodiment of the present invention.

FIG. 3 is a block diagram of an apparatus according to an embodiment of the present invention.

FIG. 4 is a part of a flowchart showing a processing example of the embodiment of the present invention.

FIG. 5 is a part of a flowchart showing a processing example of the embodiment of the present invention.

FIG. 6 is a diagram showing an example of a luminance appearance frequency distribution according to the embodiment of the present invention.

FIG. 7 is a block diagram of a conventional binarization conversion device.

[Explanation of symbols]

 401 ・ ・ ・ CPU 402 ・ ・ ・ ROM 403 ・ ・ ・ RAM 404 ・ ・ ・ Video control circuit 405 ・ ・ ・ TV signal encoder 406 ・ ・ ・ Video output DAC 407 ・ ・ ・ Monitor 408・ ・ ・ Video RAM 409 ・ ・ ・ Pointing device controller 410 ・ ・ ・ Pointing device 411 ・ ・ ・ Mobile terminal I / F unit 412 ・ ・ ・ PDA 413 ・ ・ ・ Bus 501 ・ ・ ・Input signal 502 ... Memory 503 ... Average calculation unit 504 ... Threshold value determination unit 505 ... Boundary value signal 506 ... Comparator

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] March 24, 1997

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] Detailed description of the invention

[Correction method] Change

[Correction contents]

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a binarization conversion device which performs binarization conversion from a multicolor image having density information.

[0002]

2. Description of the Related Art When an image is displayed on a black and white display unit, there is a gradation display in which the density of black and white data is changed according to the brightness level of an original color image and the image is displayed by shading.

Conventionally, various image processing methods for color still images have been proposed. However, based on the information of color image data or black-and-white multi-gradation data, black-and-white prints and images on a monochrome display. Is obtained, conversion processing is performed by various methods including binarization processing.

Therefore, in general, a certain threshold value is set, the binarization process is performed with the level as a reference, and the contour (edge) is extracted. When a document image or the like is conventionally transmitted by a device such as a facsimile, the document image or the like is divided into a plurality of pixels, and the density level of each pixel is converted by an optical means into an analog signal corresponding to each density. Image information is processed by converting an electric signal into a binary digital signal using a threshold value.

FIG. 7 is a block diagram showing a binarization process of a conventional image reading apparatus. In the figure, image data of characters and figures is input as an input signal 501 to the memory 502 as image data from a video output device (not shown) such as a video tape recorder or a personal computer.
In the memory 502, it is stored as a quantized value using, for example, an 8-bit binary number. Here, the information of the peripheral pixel density level of the pixel to be binarized is calculated by the average calculator 5
03, and the average value is calculated.

In the threshold value determining unit 504, the average calculating unit 50
The threshold value used for binarization corresponding to the average value signal calculated in 3 is output to the comparator 506. Here, in the threshold value determining unit 504, the average value calculated by comparing the average value signal calculated by the average calculating unit 503 with the boundary value signal 505 set based on the pixel density level in a predetermined range is the boundary value. When the calculated average value is smaller than the boundary value, the upper correction value is set below the pre-standardized threshold value. A threshold is used. In addition, the comparator 506
Then, the quantized value stored in the memory 502 is compared with the input threshold value, converted into a binarized signal, and output. In this way, the leap value determination unit 504 obtains a more appropriate binarized output by appropriately correcting the threshold value supplied to the comparator 506 according to the surrounding pixel density level.

[0007]

In the conventional method, when there are adjacent data whose brightness levels are close to each other,
In color image data or multi-gradation data, it was possible to distinguish by color or gradation, but in the case of a single gradation of black and white display, there is a problem that it cannot be clearly read. In addition, on such a brightness level, another threshold value may be provided at a position where adjacent data can be separated, and the data having the threshold value sandwiched between them may be inverted. There were times when it disappeared. Therefore, an object of the present invention is to provide a binarization conversion device that reduces information loss due to the binarization conversion.

[0008]

A binarization conversion device according to claim 1 is a binarization conversion device for converting luminance image information having at least luminance information into binary image information.
The brightness distribution detecting means for detecting the brightness distribution of the brightness image information, the maximum brightness level information detecting means for detecting the most brightness level data which is the brightness level showing the distribution with the most brightness distribution, and the most brightness level information are binarized. A conversion unit for converting the brightness information having the other brightness into one value of the information into the other value of the binarized information to obtain the temporary binarized image information; and edge data in the brightness of the brightness image information. First edge data generating means for generating one edge data;
Second edge data generating means for generating second edge data which is edge data in the binarized information of the binarized image information, and third edge data which is present in the first edge data and is not present in the second edge data. The third edge data generating means for generating the edge data and one value of the binarized information adjacent to the third edge data on the temporary binarized image information are used to temporarily store the third edge data in the other value. And a synthesizing means for synthesizing the binarized image information. The binarization conversion device according to claim 2 is the binarization conversion device according to claim 1, wherein the third edge data generation means is an exclusive OR of the first edge data and the second edge data. Is obtained to generate the third edge data, and the synthesizing means synthesizes the temporary binary image information and the third edge data by taking an exclusive OR.

[0009]

In the binarization conversion device according to the present invention, the maximum luminance level information is converted into one value of the binarized information, and the luminance information having the other brightness is converted into the other value of the binarized information, and the temporary 2 It is possible to obtain the binarized image information, prevent the information from being lost by a plurality of binarization processes, and invert and display the binarized image having a large area.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described with reference to FIGS.
This will be described with reference to FIG. 1 and 2 are examples showing a method of an embodiment of the present invention. FIG. 3 is a block diagram showing the configuration of a control device that implements and operates software for performing the processing of the present invention. In FIG. 3, 401
Is a CPU (Central Processing U)
ROM (Read) via the bus 413.
Only Memory 402, RAM (Rand)
om Access Memory (403) to control the peripheral devices 404, 408, 409, 411. The computer program of the software is stored in the ROM 402.

The image data to be processed is obtained from a CD-ROM, a floppy disk or a communication line (not shown), and the processed image data is stored in the portable terminal I / F section 41.
1 through PDA (Personal Digital
l Assistant) 412 and other portable terminals. The captured image data is captured in the video RAM 408. The main operation is performed from the pointing device 410 such as a mouse through the pointing device control unit 409.

When the image signal of FIG. 1A is a color signal, the CPU 401 re-stores in the video RAM 408 the brightness conversion image converted from the color signal to the brightness signal by the software of the ROM 402. Next, measure the appearance frequency with respect to the brightness of this brightness data with software,
The luminance distribution as shown in FIG. 1 (b) is examined. The analysis result is stored in the RAM 403 as peak brightness.

Next, the maximum luminance among the luminances stored in the RAM 403 is selected as the reference value ref, and the difference between the luminance of the reference value and the adjacent peak luminance and the luminance of the reference value is selected. Is calculated by software, RAM4
03. This luminance difference is 2 as shown in FIG.
Used as the range rng for quantification.

Next, the brightness data of FIG. 1A is binarized and converted by software using the above-mentioned reference value ref and range rng to obtain FIG. That is, the brightness is re
The data within f ± rng / 2 is stored in another frame of the video RAM 408 as 0 (black) and the other data as 1 (white). In the example of FIG. 1C, as can be seen from FIG. 1D, only the blue color of the sea is 0 (black) and the other portions are 1 (white).

Next, the image of FIG. 1E is obtained from the binarized image data by edge detection (outline extraction). That is, the boundary between the sea and other parts is obtained as a contour. Edge detection is performed as follows, for example. The edge refers to a boundary where areas of different colors are adjacent to each other. Therefore, if the difference between the image luminance data in one horizontal scanning period and the data obtained by delaying the data by one pixel or several pixels is obtained, the luminance change data in the horizontal scanning period can be obtained.

For the vertical scanning period, the brightness change data in the vertical direction can be obtained in the same manner by using a signal delayed by one horizontal scanning period or several horizontal scanning periods. Here, this processing is also performed by software. Ie video RAM
Software processing is performed from the binarized image data accumulated in 408, and stored in another frame of the video RAM 408.

A computer program is stored in the ROM 402, and the CPU 401 executes the software processing by using the RAM 403 and the video RAM 408 according to the program of the ROM 402 via the bus 413.

Then, the brightness-converted image to be processed is edge-detected in the same manner to obtain the image data of FIG. 1 (f). This is also done by software processing, and the video RAM 40
8 in yet another frame. This Figure 1 (f)
Means that the contour of FIG. 1A has been extracted.

Next, FIG. 1 stored in the video RAM 408.
If the exclusive OR of the image data of (e) and the image data of FIG. 1 (f) is performed by the software processing, as shown in FIG.
Data existing only in one of (e) and FIG. 1 (f) can be extracted.

Finally, the exclusive OR of the binarized image data of FIG. 1 (c) and the edge data of FIG. 2 (g) is software processed to obtain the image data of FIG. 2 (h).
That is, in the image of FIG. 2H, the road portion is displayed in black on the land side and white on the sea side in the binarized image of black of the sea and white of the land, and necessary information is obtained as a binarized image. be able to.

In this way, the CPU 401 controls the ROM 4
02 software, the image data accumulated in the video RAM 408 is binarized through the RAM 403, the result is accumulated in the RAM 403, and the TV signal is transmitted through the video control circuit 404 according to the instruction of the pointing device 410. Output to encoder 405,
The video output DAC 406 outputs the analog signal to the monitor 407.

Here, FIG. 1B shows the information of FIG. 1A with the horizontal axis representing the luminance level and the vertical axis representing the appearance frequency. For example, maps are generally represented by outlines of symbols that represent roads, railways, stations, contour lines, sea lakes, rivers, etc., characters that show proper nouns of each, and color divisions that represent cities, towns, wards, towns, and their boundaries. ing. In the case of FIG. 1 (a),
Shown is the case where the sea is shown in blue, the road is shown in gray, the land is shown in light brown, and it is displayed in black letters.

Characters, boundaries, railroads, main roads, etc. are often expressed in dark black, red, blue, etc., general roads are gray, sea lakes and rivers are light blue, park green areas are light green, etc. expressed.
In addition, buildings and the like surrounded by contour lines are often painted in different colors to distinguish them from each other. Therefore, for example, the luminance-frequency distribution of the map is obtained as shown in FIG. In addition, when the input data is obtained from a commercial map or the like with a scanner or the like, for example, there may be a case where the originally existing information such as dust or added data is not included.

However, roughly speaking, if the above description is made, the part having the highest brightness corresponds to white, the part having the lowest brightness corresponds to black, and the black part has a contour line or a boundary line and character information. The gray area is the general road, and the white area is the distribution map of the color area representing the area division such as other rivers, lakes and rivers.

[0025] Here, contour lines are used for most boundaries in normal maps, but examples are considered such as sea-land boundaries where contour lines are omitted in peripheral maps of road maps and tourist maps. To be In such a case, for example, if a blue portion of the sea or the like and a light brown portion of the land are discriminated and binarized, portions lower than the blue and blue luminance levels are represented by black, and portions other than that are represented by white. . Then, in the sea filled with blue, all the parts such as bridges and letters in which the brightness is lower are black, and the information is lost.

In such a case, the luminance as a color-
In the frequency distribution, frequencies of blue, light brown, gray, and black appear in order of frequency. The order of brightness levels is light brown, blue, gray, and black. Therefore, FIG. 1C shows that the brightness level of blue is binarized to black and the others are binarized to white. This is the temporary binarized image information, and the luminance-frequency distribution before the binarization is shown in FIG. As shown in FIG. 1D, the range in which the black level is binarized is a range in which the difference between the brightness of blue and the brightness of ash adjacent to blue is centered. Therefore, in this case, the range of the difference in brightness between blue and grey, centered on blue, is black, and only blue is replaced with black and the others are replaced with white.

FIG. 1E shows the result of edge detection from FIG. 1C, and FIG. 1E shows the result of edge detection from FIG. 1A of the display example to be converted. f). Then, the roads and characters missing in FIG. 1E appear in FIG. 1F. Therefore, if the exclusive OR of FIG. 1 (e) and FIG. 1 (f) is taken, the information in either FIG. 1 (e) or FIG. 1 (f) is obtained as FIG. 2 (g). To be Therefore, if the exclusive OR of FIG. 2 (g) and FIG. 1 (c) of the previous temporary binarized image information is taken, roads and characters in the land portion are added to FIG. 1 (c). Figure 2
(H) is obtained.

FIG. 6 shows an example of the above-mentioned luminance-frequency distribution. FIG. 6A shows the case of a yellow single color screen.
FIG. 6B is a diagram schematically showing a case of black characters or a black frame on a white background, and FIG. 6C is a diagram schematically showing a case of multicolor complicated display. Each,
The frequency peak numbers are 1, 2, and 10.

Next, an example of a flowchart showing the binarization processing method of the present invention is shown in FIGS. STEP1
Brightness distribution detection in STEP 5, STEP 6 to STEP 13
In step 14, the second edge data is generated in STEP 14, the first edge data is generated in STEP 15, and the third edge data is generated in STEP 16.
At 17, the third edge data and the temporary binarized image information are combined.

From STEP 1 start, first, STEP 2
Then, one screen of image data to be processed, for example, the data shown in FIG. 1A is loaded into the video RAM 408, and brightness data conversion is performed in STEP3. The luminance data conversion is performed, for example, when the image data to be processed is RGB data, Y = 0.2988 × R + 0.5868 × G +
The luminance component Y is obtained from 0.5868B, and in the case of YUV data, Y
Use only data.

The brightness data thus obtained is A. Next, in STEP 4, the luminance distribution analysis is performed from the luminance data for one screen. That is, the luminance is plotted on the horizontal axis and the frequency of appearance on the vertical axis, and the luminance distribution for one screen is examined. This STEP corresponds to FIG. In STEP 5, Y ₁ (Pmax) having the highest frequency is set as the reference value ref from the result of the luminance distribution analysis (FIG. 1 (d)).

In STEP 6, the number of local peak values of frequency is checked from the result of the luminance distribution analysis, and it is checked whether the number is 2 or more. If Yes, the process proceeds to the setting of the range rng in STEP 7, and the difference between the highest frequency luminance Y ₁ and the frequency Y _{2 of the} frequency peak closest to Y ₁ on the luminance axis is set as the range rng. For example, in the case of FIG. 1 (d),
Since Y ₁ corresponds to the brightness of blue in the sea and the brightness of ash corresponds to Y ₂ , the difference between the brightness levels of blue and ash is set as the range rng. On the other hand, when the result in STEP 6 is No, the frequency peak is 1, which corresponds to the case where the entire screen is composed of a single color. In this case, the difference between the highest luminance and the lowest luminance is set as the range rng in STEP8.

Next, in STEP 9, the luminance of the reference value Y ₁
Is the central brightness, that is, the difference between the maximum brightness and the minimum brightness is 1.
Check if it is higher than / 2, and if Yes,
The parameter P used later is set to 0 (STEP1
0). If No, P is set to 1 in STEP 11.

Next, in STEP 12, using the reference value ref, the range rng, and the parameter P set in the above, each data of the brightness data A is converted into the brightness data Y ₁ ±.
If it is within rng / 2, it is set to P, and the other data is set to the inverted value of P. The image data thus obtained is called binarized data B. This binary data B is temporary 2
It becomes the digitized image information.

At STEP 13, it is checked whether the number of peaks having a frequency equal to or higher than the specified value is 2 or more. The specified value is set, for example, 70% or more of the frequency of the maximum brightness, or the like, from the visual effect. If Yes, in STEP 14, edge detection is performed from the binarized data B, and the binarized data C is detected.
And This binarized data C becomes the second edge data. In FIG. 1, it corresponds to (e).

In STEP 15, edge detection is performed from the brightness data A to obtain binarized data D. This is shown in FIG.
Corresponding to (f), the binarized data D becomes the first edge data. Further, in STEP 16, the exclusive OR (EXOR) processing of the binarized data C and the binarized data D is performed, and the obtained data is set as the binarized data E. Binary data E
Is the third edge data and corresponds to FIG. Further, in STEP 17, the exclusive OR (EXOR) processing of the binarized data B and the binarized data E is performed, and the obtained data is set as the binarized data F. In STEP 17, the third edge data and the temporary binarized image information are combined, and the resulting binarized data F corresponds to FIG.

As a result, the binarized data is obtained by exclusive OR processing of the edge data of the binarized data C that has undergone the luminance range processing and the binarized data D of the edge data obtained from the original luminance data. Edge data that does not appear in D is obtained as binarized data E, and the desired binarized data F is obtained by exclusive OR processing of the data with the binarized data B that is the binarized data of the area. Will be obtained.

STEP 18 shows the case of No in STEP 13, which is the case where the number of peaks having the brightness equal to or higher than the specified value is 1 or 0. In this case, the binarized data B is output. After STEP 19, it is whether or not the output is inverted output. Depending on the setting or instruction, select whether to output as it is or invert output in STEP 19, and if it is inverted output, STEP 20; if it is output as it is, flow through STEP 21. Ends and the end of STEP 22 is reached.

As described above, the process has been described as a process for one screen, but a masking process is performed as a part of the process, that is, a partial region of the image is designated and the above process is partially performed and fitted. It is also possible to obtain a desired image by performing. Further, even if the image to be processed is a binary image or another format such as vector processing data, it is converted into a format usable for the main conversion processing by the pre-processing and converted into another format by the post-processing. By doing so, it can be applied to image data of other formats.

[0040]

As described above, in the binarization conversion device of the present invention, the loss of information due to the binarization conversion can be reduced, and the binary data area having a large area can be displayed inverted. After conversion, it is possible to obtain a more effective display visually.

Claims (2)

[Claims] 1. A binarization conversion device for converting luminance image information having at least luminance information into binarized image information, the luminance distribution detecting unit detecting a luminance distribution of the luminance image information, and the luminance distribution. The maximum brightness level information detecting means for detecting the maximum brightness level data which is the brightness level showing the most distribution of the distribution, the maximum brightness level information as one value of the binarized information, and the brightness information having the other brightness as 2 Conversion means for converting into the other value of the binarization information to obtain temporary binarized image information; first edge data generation means for generating first edge data which is edge data at the luminance of the luminance image information; Second edge data generating means for generating second edge data which is edge data in the binary information of the temporary binary image information; and the second edge data existing in the first edge data. Third edge data generating means for generating third edge data that is edge data that does not exist in the data, and one value of the binarized information adjacent to the third edge data on the temporary binarized image information. , A synthesizing means for synthesizing the third edge data with the other value on the temporary binarized image information. 2. The third edge data generating means generates third edge data by taking an exclusive OR of the first edge data and the second edge data, and the synthesizing means generates the temporary binary value. 2. The binarization conversion device according to claim 1, wherein the binarized conversion device is configured to combine the binarized image information and the third edge data by taking an exclusive OR.