JP3679619B2 - Image contour extraction method, apparatus, and recording medium - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image contour extraction method and apparatus, and a recording medium on which a program for extracting an image contour by a computer is recorded.
[0002]
[Prior art]
When an image is processed by a computer, an outline of the image may be extracted. The contour of the extracted image is subjected to conversion processing such as rotation, enlargement, reduction, etc. on the image in the contour, the image in the contour is filled with a desired color, the image in the contour is combined with other images, or descending It is used for processing.
[0003]
Here, as a method for extracting the contour of an image, there is a method disclosed in Japanese Patent Laid-Open No. 5-242246. In this method, the image is scanned line by line along the main scanning direction, and a contour line along the main scanning direction is extracted in each scanning. If the end points of the two contour lines detected in the main scanning direction match, they are connected in the sub scanning direction to form one contour line. A desired contour is extracted by performing such processing for each line of all images.
[0004]
[Problems to be solved by the invention]
However, in the conventional image contour extraction method as described above, when several contour lines with different density values are extracted, the entire image region is scanned line by line to extract one contour line. Processing had to be performed for each density value, and there was a problem that the time required for contour extraction was long.
[0005]
The present invention has been made in view of the above problem, and records an image contour extraction method, apparatus, and program for efficiently performing contour extraction processing for an arbitrary density value and reducing processing time. An object of the present invention is to provide a recording medium.
[0006]
[Means for Solving the Problems]
In order to achieve the above object, the invention according to claim 1 is a method for extracting an outline of an image, wherein: (a) a pixel of interest one by one in descending order from a pixel indicating the maximum density value in the image And (b) when a plurality of contour sides around the target pixel are not included in the existing contour line, a contour line composed of the plurality of contour sides is set as a new contour line, and Contour line connection information indicating a connection state of a new contour line is generated, and the start side of the new contour line selected from any one of the plurality of contour sides of the target pixel, and the density of the target pixel Generating contour management information including a starting density value whose value is a starting density of the effective density range of the new contour, and (c) any one of the plurality of contour sides of the target pixel is already formed. If included in the contour line, Updating the outline connection information for a line; and (d) extinguishing if at least one existing outline that includes any of the outline sides of the pixel of interest disappears due to the updating. Adding a density value of the pixel of interest as an end density value indicating an end density of an effective density range of the existing contour line to the contour management information for the existing contour line.
[0007]
According to a second aspect of the present invention, in the image contour extracting method according to the first aspect, the steps (a) to (d) are repeatedly performed by using a pixel having the same density value in the image as the target pixel. The contour line connection information generated / updated by this is output.
[0008]
According to a third aspect of the present invention, in the image contour extracting method according to the first aspect, (e) storing the contour management information generated by repeatedly performing the steps (a) to (d). And when extracting a contour line of a predetermined density value, (f) the predetermined density based on the effective density range indicated by the start density value and the end density value included in the contour management information Specifying a contour line of density value, obtaining the start side from the contour management information for the contour line, and (g) sequentially extracting contour sides from the start side according to a predetermined rule; It is characterized by having.
[0009]
According to a fourth aspect of the present invention, in the image contour extracting method according to any one of the first to third aspects, the step performed prior to the steps (a) to (d) includes: (h) the image (I) a step of securing a recording area for recording pixel information for all the pixels constituting the image, and (i) searching for all the pixels constituting the image and counting the number of pixels for each density value, and (j ) Obtaining a cumulative number of pixels less than the density value for each density value based on the count result; (k) extracting the pixels constituting the image pixel by pixel; and (l) the ( k) detecting the density value of the pixel extracted in the step, obtaining the cumulative pixel number corresponding to the density value, and extracting the extracted pixel at the position of the recording area corresponding to the cumulative pixel number value. The pixel information to be specified is recorded, and the value of the cumulative number of pixels And a step of updating is characterized repetition by performing said (l) step from (k) until all of the pixels constituting the image.
[0010]
The invention according to claim 5 is an apparatus for extracting a contour of an image, and (a) a target pixel extracting unit that extracts a target pixel one by one in descending order from a pixel indicating a maximum density value in the image; (B) When a plurality of contour sides around the target pixel are not included in the existing contour line, the contour line composed of the plurality of contour sides is set as a new contour line, and the connection state of the new contour line is And generating a contour connection information to indicate the start side of the new contour line selected from any one of the plurality of contour sides of the target pixel and the density value of the target pixel of the new contour line. Contour line information generating means for generating contour line management information including a start density value as the start density of the effective density range; (c) storage means for storing the contour line connection information and the contour line management information; (d) Of the plurality of contour sides of the target pixel If any of the contours is included in the existing contour, the contour connection information stored in the storage unit is updated for the existing contour, and the contour of any of the target pixels is updated by the update. In a case where at least one existing contour line that includes an edge disappears, the density value of the target pixel is set with respect to the contour management information stored in the storage unit for the existing contour line that disappears. Updating means for additionally updating as an end density value indicating the end density of the effective density range of the existing contour line.
[0011]
The invention according to claim 6 is a recording medium on which a program for extracting the contour of an image is recorded by a computer, and (a) 1 in descending order from a pixel indicating the maximum density value in the image. A procedure for extracting a pixel of interest for each pixel; (b) when a plurality of contour sides around the pixel of interest are not included in the existing contour line, a contour line composed of the plurality of contour sides is set as a new contour line And generating the contour connection information indicating the connection state of the new contour, and selecting the start side of the new contour that selects any one of the plurality of contour sides of the pixel of interest, and the attention (C) any one of the plurality of contour sides of the target pixel; and (c) a procedure for generating contour management information including a density value of a pixel and a start density value that is a start density of an effective density range of the new contour line. Is an existing outline If included, a procedure for updating the contour connection information for the ready contour, (d) at least one ready contour in which any of the contour sides of the target pixel is included by the update When the line disappears, the density value of the target pixel is added as the end density value indicating the end density of the effective density range of the ready contour line to the contour management information regarding the ready contour line that disappears. An image contour extraction program for executing the procedure is recorded.
[0012]
DETAILED DESCRIPTION OF THE INVENTION
<1. Summary of Invention>
The above-described conventional image contour extraction method scans the contour extraction process for each line in the sub-scanning direction along the main scanning direction on the image plane composed of the main scanning direction and the sub-scanning direction.
[0013]
On the other hand, the image contour extraction method according to the present invention sequentially extracts one pixel at a time from the pixel showing the maximum density value out of the pixels constituting the image, and generates a new contour line, Or update the line.
[0014]
Here, if a line that is a boundary between one pixel and its adjacent pixels is called a contour side, one pixel generally has four contour sides.
[0015]
Therefore, it is determined whether or not any of the four contour sides of the pixel sequentially extracted from the maximum density value (hereinafter referred to as “target pixel”) is included in the already generated contour line. If none of the contour sides is included in the existing contour line, a new contour line consisting of the four contour sides of the target pixel is newly generated, while one of the contour sides is included in the existing contour line. In this case, the existing contour line is updated so that the target pixel is included in the contour.
[0016]
FIG. 1 is a diagram showing the concept of such contour generation / update. 1 exemplifies 2 × 2 four pixels PA to PD, the density value of the pixel PA is LV4, the density value of the pixel PB is LV3, and the density value of the pixel PC is LV2. The density value of PD is LV1. Here, the relationship between the density values is LV1 <LV2 <LV3 <LV4. First, when the maximum density value LV4 is extracted, the pixel PA is extracted. Then, as shown in FIG. 1A, the four contour sides of the pixel PA are generated as the new contour line Ea. Next, as shown in FIG. 1B, a pixel PB having a density value LV3 is extracted. Of the four contour sides of the pixel PB, the contour side in contact with the pixel PA is already included in the contour line Ea, so the contour line Ea is updated to generate the contour line Eb. Next, as shown in FIG. 1C, the pixel PC having the density value LV2 is extracted. Of the four contour sides of the pixel PC, the contour side in contact with the pixel PA is already included in the contour line Eb, so the contour line Eb is updated to generate the contour line Ec. Finally, as shown in FIG. 1D, the pixel PD having the density value LV1 is extracted. Of the four contour sides of the pixel PD, two contour sides in contact with the pixels PB and PC are already included in the contour line Ec, so the contour line Ec is updated to generate the contour line Ed.
[0017]
In this process, all the contour lines Ea to Ed existing for each density value are extracted while the four pixels PA to PD are processed one by one. Therefore, if the contour line is output when the contour line for each density value is extracted, the contour line for all density values can be obtained by performing the process for each pixel for the entire image region only once. Can be obtained.
[0018]
Further, when a new contour line is generated in such a process, one of the contour sides constituting the new contour line is set as a start side so that the contour line can be traced later. Set as. For this reason, even if the contour line is updated as the above process proceeds, it is possible to easily extract a contour line having a predetermined density value by tracking the contour based on a predetermined rule (rule) from the start side. it can.
[0019]
Hereinafter, a detailed configuration for realizing such processing will be described.
[0020]
<2. Configuration of device>
FIG. 2 is a block diagram showing a configuration of an image contour extracting apparatus 100 as an embodiment of the present invention. As shown in FIG. 2, the contour extracting apparatus 100 includes a preprocessing unit 10, a target pixel extracting unit 20, a contour side determining unit 30, a contour line forming unit 40, a memory 50, a contour extracting unit 60, and a storage unit 70. I have.
[0021]
The storage unit 70 is for storing an image as data, and the preprocessing unit 10 is a part that reads an image from the storage unit 70 and performs preprocessing on the image. In the preprocessing, the generation of an image with a frame, the assignment of contour edge numbers, and the sorting of pixels based on density values are performed. Details of these processes will be described later.
[0022]
The pixel-of-interest extraction unit 20 sequentially extracts pixels of interest one pixel at a time in order from the pixel having the highest density value among a plurality of pixels constituting the image.
[0023]
Then, the contour edge determination unit 30 determines whether or not any of the four contour edges of the target pixel is included in the already generated contour line, and the result is sent to the contour line formation unit 40. Tell. In determining the contour side, the contour connection information DC stored in the memory 50 is referred to. The contour line is formed by continuously connecting a plurality of contour sides. Since the connection state of the plurality of contour sides forming the contour line has already been described in the contour line connection information DC, the contour line connection information By referring to DC, it can be determined whether or not any of the four contour sides of the target pixel is included in the already generated contour line.
[0024]
FIG. 3 is a diagram for explaining the outline connection information DC. For example, as shown in FIG. 3A, when a pixel PA is extracted as a target pixel and a new contour line is formed by four contour sides a, b, c, d around the pixel PA. The outline connection information DC is described in a state as shown in FIG. 3B, the contour side a is connected to the contour side b, the contour side b is connected to the contour side c, the contour side c is connected to the contour side d, and the contour side d is connected to the contour side a. You can see that The outline number of the outline formed by these outline sides a, b, c, and d is “1”.
[0025]
Then, when the pixel PB shown in FIG. 3A is extracted as the target pixel, if the contour line connection information DC in FIG. 3B is referred to, the four contour sides b, e, f around the pixel PB will be described. , G, the contour side b is already included in the contour line with the contour number “1”. Therefore, the use state of each contour side of the target pixel can be determined by referring to the contour line connection information DC. One contour side cannot be included in a plurality of contour lines at the same time.
[0026]
The contour line forming unit 40 includes a contour line information generating unit 40a and a contour line information updating unit 40b, and the contour side determining unit 30 has already generated all of the four contour sides of the target pixel. When it is determined that the line is not included in the line, the outline information generation unit 40a operates to generate a new outline, while one of the four outline sides of the target pixel has already been generated. When it is determined that it is included in the existing contour line, the contour line information update unit 40b operates to update the existing contour line. This update processing includes processing of contour separation, integration, and disappearance as will be described later.
[0027]
When the contour information generation unit 40 a generates a new contour, the contour connection state is described in the contour connection information DC, and the contour management information DM is generated and stored in the memory 50.
[0028]
The outline information update unit 40b updates the connection state of the outline described in the outline connection information DC when the existing outline is updated.
[0029]
Further, when at least one of the existing contour lines disappears due to the update, the contour line information update unit 40b stores the density value of the current target pixel in the end density value column of the contour line management information DM for the disappeared contour line. One higher density value is stored. Here, when at least one of the existing contour lines disappears due to the update, the start side registered in the contour management information DM becomes invalid due to the addition of the pixel of interest, and two or more contour lines This includes the case where only one contour line remains due to the integration and the other contour line disappears. Further, when there is no pixel of interest to be extracted next, the contour line information update unit 40b stores “0” as the end density value for those for which the end density value is not yet stored in the contour line management information DM. To do. Note that processing for storing the end density value in this way is referred to as end processing.
[0030]
Since the above-mentioned contour line connection information DC is the contour line itself, the contour line information generation unit 40a and the contour line information update unit 40b describe the new connection state in the contour line connection information DC or the contour line connection information DC. If the contour line connection information DC is output when updated, the contour line is output.
[0031]
For example, when the density value of the image is 256 gradations from 0 to 255 and it is desired to extract a contour line having a density value of 126, the pixel of interest is extracted sequentially from the pixel of the density value 255, and each of them is extracted. Contour lines for density values are formed. When the processing for all the pixels having a density value of 126 is completed, a desired contour line has been formed in the contour line connection information DC. In this case, extraction of pixels having a density value of 125 or less need not be performed, so efficient contour extraction is performed.
[0032]
However, in this embodiment, once the contour line for each pixel is formed for all the images, it is necessary to form the contour line for each pixel for all the images when trying to extract the contour line again. In order to avoid this, the contour management information DM is stored in the memory 50.
[0033]
Here, the outline management information DM will be described. FIG. 4 is a diagram for explaining the outline management information DM. The numbers in each pixel shown in FIG. 4A indicate the density value of each pixel.
[0034]
For example, as shown in FIG. 4A, when a pixel PA is extracted as a target pixel and a new contour line is formed by four contour sides a, b, c, and d around the pixel PA. The outline management information DM is described in a state as shown in FIG. A number that does not overlap with an existing contour number is assigned to the contour number of the contour management information DM. The start side number stores the number of the contour side (that is, the start side) that becomes the extraction start point when a newly generated contour line is extracted later.
[0035]
As a start side, a contour side shared with a pixel indicating the minimum density value among pixels adjacent to the target pixel, or a density value having the largest difference from the density value of the target pixel among pixels adjacent to the target pixel It is preferable to select a contour side shared with a pixel having. This is because the effective density range in which the start side is effective can be maximized. Therefore, in the case of the density value as shown in FIG. 4A, the contour side a is determined as the start side.
[0036]
In the start density value, the density value “9” of the target pixel PA is stored. The end density value stores a density value that is one higher than the density value of the pixel of interest when the outline with the outline number “1” is updated and disappears, that is, when the start side becomes invalid. .
[0037]
In the number of contour sides, the number of contour sides constituting the contour line of the contour number “1” is stored. In the case of the example of FIG. 4A, the contour line is composed of four contour sides a, b, c, and d, so the number of contour sides is “4”.
[0038]
By storing such contour management information DM in the memory 50, when a contour line having an arbitrary density value is extracted later, the contour extraction unit 60 starts and ends the density values of the contour management information DM. The effective density range in which the start side of the contour line is effective can be known from the value, and all the start sides in which the density value to be subjected to the contour extraction is within the effective density range can be acquired. Then, if the contour extraction unit 60 performs contour extraction based on a predetermined rule from those start sides, it is possible to appropriately extract the contour of the image.
[0039]
That is, in the apparatus of FIG. 2, when the contour management information DM is not yet stored in the memory 50, the pre-processing unit 10, the target pixel extracting unit 20, the contour side determining unit 30, and the contour forming unit 40 are included. It functions so that the contour line management information DM and the contour line connection information DC are sequentially stored in the memory 50 as the contour line for each density value is formed. Further, the contour line forming unit 40 can perform contour line output by outputting the contour line connection information DC when it is generated / updated.
[0040]
On the other hand, when the outline management information DM is already stored in the memory 50, the outline extraction unit 60 functions to read out the outline management information DM in the memory 50 and register it for each outline. Contour extraction is performed from the side based on a predetermined rule, and the contour line extracted by the contour extraction unit 60 is output.
[0041]
<3. Pretreatment>
The preprocessing performed in the preprocessing unit 10 will be described.
[0042]
First, a framed image is generated as first preprocessing. FIG. 5 is a diagram illustrating generation of a framed image. FIG. 5A shows an original image to be subjected to contour extraction, FIG. 5B shows an intermediate stage when generating a framed image, and FIG. 5C shows the original image. The image with a frame produced | generated is shown. FIG. 5 shows an example in which the original image is composed of 5 × 4 20 pixels, and the number in each pixel indicates the density value of that pixel. In the following, the process for the original image as shown in FIG. 5A will be described as an example.
[0043]
The preprocessing unit 10 secures an image area in which the periphery of the original image in FIG. 5A is expanded by one pixel (FIG. 5B). As a result, the density value 0 is stored for all the pixels distributed around the original image. In this way, a framed image made up of 42 pixels of 7 × 6 is generated as shown in FIG.
[0044]
The reason for generating such a framed image is to make it possible to perform the same processing on all the contour sides of the original image and improve the processing efficiency. Since each contour side located at the outer periphery of the original image shown in FIG. 5A is not in contact with other pixels, the density value comparison cannot be performed. Therefore, a framed image is generated so that density value comparison can be performed for all contour sides included in the original image. Note that it is not necessary to generate a framed image when exception processing different from other contour sides is performed for each contour side located on the outer periphery of the original image shown in FIG.
[0045]
Next, as a second preprocessing, a contour side number assignment process is performed for each contour side including the framed image. FIG. 6 is a diagram illustrating the assignment of contour side numbers. If there are n pixels in the main scanning direction X in the framed image, 0 is assigned to the lower side of the leftmost pixel of the first line, and 1 is assigned to the right side. Then, numbers are sequentially assigned to the lower side, the right side, the lower side of the third pixel, and the right side of the second pixel in the first line. Since there are n pixels in the main scanning direction X, the number (2n) is assigned to the lower side of the leftmost pixel of the second line, and the number (2n + 1) is assigned to the right side. In the same manner, numbers are sequentially assigned to the lower side, the right side, the lower side, and the right side of the second pixel in the second line of the second line, and contour side numbers are assigned to all lines.
[0046]
In this manner, when the contour side numbers are assigned to the 7 × 6 framed image as shown in FIG. 5C, the result is as shown in FIG. In FIG. 7, the thick line indicates the original image portion. Therefore, by assigning the contour side number, a unique number is assigned to all the contour sides necessary for the contour extraction of the original image. Here, as shown in FIG. 7, no contour side number is assigned to each contour side located at the upper end of the framed image and each contour side located at the left end, but these contour sides are the contours of the original image. There is no problem because it is impossible.
[0047]
Also, when assigning contour edge numbers only to the original image, time must be taken because exception processing must be performed to assign contour edge numbers to the contour edges located at the top and left edges of the original image. However, when assigning contour edge numbers to an image with a frame as described above, numbers are assigned to all contour edges included in the original image without performing exception processing. The contour side number can be given to the processing time, and the processing time can be shortened.
[0048]
Next, as a third preprocess, a process of sorting the pixels constituting the original image based on the density value is performed. The sorting based on the density value is performed so that the target pixel extraction unit 20 can sequentially extract the target pixels from the pixels having the higher density values. Accordingly, any known sorting method can be applied as long as it can be sorted so that one pixel can be extracted in descending order of density value from the pixels constituting the original image. can do.
[0049]
However, the various sort methods known in the art require a lot of time for sorting because judgment processing and data replacement processing are frequently used.
[0050]
Therefore, in this embodiment, it is preferable to apply a sorting method as described below in order to perform sorting efficiently.
[0051]
First, prior to performing the sorting process, a coordinate value is set for each pixel of the framed image. In general, when coordinate values are set for each pixel in the image plane, an XY coordinate system expressed by a position in the main scanning direction X and a position in the sub-scanning direction Y is used. In this case, the X coordinate and the Y coordinate are used. The two values of must be managed. Therefore, in this embodiment, if the framed image has n pixels in the main scanning direction X, one value derived by “(Y coordinate) × (n) + (X coordinate)” is expressed as coordinates. Value. That is, a coordinate value is set for each pixel by sequentially assigning one number along the main scanning direction X of the framed image. When the coordinate values are set for the framed image shown in FIG. 5C in this way, the coordinate values of each pixel are as shown in FIG. In FIG. 8, the number of each pixel is a coordinate value, and the display of the density value is omitted. Further, the thick line in FIG. 8 indicates the original image portion.
[0052]
Next, the actual sorting process is performed. First, all pixels included in the original image are searched, and the number of pixels for each density value is counted. When the number of pixels for each density value is counted for the original image shown in FIG. 5, a count result as shown in FIG. 9 is obtained. The preprocessing unit 10 temporarily stores the count result in a memory (not shown) provided therein.
[0053]
Then, based on the count result shown in FIG. 9, for each density value, the cumulative number of pixels indicating a density value less than the density value is derived. For example, from the count result of FIG. 9, the cumulative number of pixels with density value 1 is 0 because it is the number of pixels with density value 0, and the cumulative number of pixels with density value 2 is the sum of the number of pixels with density values 0 to 1. Therefore, the number of accumulated pixels having a density value of 3 is 5 because it is the sum of the number of pixels having density values of 0 to 2. When the cumulative number of pixels is obtained for each density value in this way, it is as shown in FIG.
[0054]
Next, buffer areas for storing sort data are secured for the number of pixels constituting the original image. Since the original image shown in FIG. 5 has 20 pixels, 20 buffer areas are secured. Then, one pixel is extracted from the pixels constituting the original image in the order of coordinate values, and the density value is detected.
[0055]
In the case of the original image in FIG. 5, pixels with a density value of 2 are first extracted as shown in FIG. 11. Since the cumulative number of pixels corresponding to the extracted pixel density value 2 is 2, the coordinate value of the extracted pixel is stored in the storage position corresponding to 2 in the buffer area, that is, the memory address 2 position. In the case of FIG. 11, since the coordinate value of the pixel having the extracted density value 2 is 8 (see FIG. 8), the coordinate value 8 is stored at the position of the memory address 2. When the coordinate value is stored in the buffer area, the cumulative number of pixels of density value 2 is changed to 3 by adding 1 to the cumulative number of pixels 2 of density value 2.
[0056]
Next, as shown in FIG. 12, a pixel having a density value of 9 is extracted. Since the cumulative number of pixels corresponding to the extracted pixel density value 9 is 19, the coordinate value of the extracted pixel is stored in the storage position corresponding to 19 in the buffer area, that is, the memory address 19 position. In the case of FIG. 12, since the coordinate value of the pixel of the extracted density value 9 is 9 (see FIG. 8), the coordinate value 9 is stored at the memory address 19 position. When the coordinate value is stored in the buffer area, 1 is added to the accumulated pixel number 19 of the density value 9 to change the accumulated pixel number of the density value 9 to 20.
[0057]
Thereafter, the same processing is repeated every time one pixel is arranged in the coordinate value order from the pixels constituting the original image. When the processing for all the pixels constituting the original image is completed, the cumulative number of pixels for each density value and the buffer area are in a state as shown in FIG.
[0058]
In the buffer area shown in FIG. 13B, the coordinate values of the pixels having the smaller density values are stored in order from the smallest memory address. For example, the memory addresses 0 and 1 store the coordinate value of the pixel having the density value 1, and the memory addresses 2 to 4 store the coordinate value of the pixel having the density value 2. Therefore, by performing the processing as described above, the result of sorting the pixels constituting the original image based on the density value is stored in the buffer area.
[0059]
When the sorting method as described above is applied, the sorting time can be reduced without performing the determination process for comparing the density values of the two pixels or the data replacement process, thereby reducing the time required for sorting.
[0060]
If the sort result based on the density value is stored in the buffer area in this way, when the target pixel extracting unit 20 sequentially extracts the target pixel from the pixels having the higher density value, the maximum memory address of the buffer area is stored. It is only necessary to acquire coordinate values sequentially from the positions and extract pixels having the coordinate values, and efficient processing can be performed.
[0061]
<4. Generation and update of contour management information>
Next, a description will be given of processing in which the target pixel extraction unit 20 sequentially extracts target pixels from pixels having a high density value, and generates or updates the contour management information and the contour connection information.
[0062]
The target pixel extraction unit 20 sequentially extracts target pixels from pixels having a high density value based on the sort result as shown in FIG. Then, the contour side numbers of the four contour sides around the target pixel are obtained. As shown in FIG. 13B, since the coordinate value of the pixel is stored in the sort result, the numbers of the four contour sides of the target pixel are obtained from this coordinate value. For example, if the coordinate value of the pixel of interest is P, and n pixels exist in the main scanning direction X of the framed image, the contour edge numbers on the upper side, the right side, the lower side, and the left side of the pixel of interest are NU, If NR, ND, NL,
Upper contour side number: NU = (P−n) × 2
Contour side number on the right side: NR = (P × 2) +1
Lower contour side number: ND = (P × 2)
Contour side number on the left side: NL = (P × 2) −1
Can be obtained. Then, the target pixel extraction unit 20 sends the four contour side numbers for the target pixel to the contour side determination unit 30.
[0063]
The contour side determination unit 30 refers to the contour line connection information DC in the memory 50 and determines whether or not the contour side number of the target pixel is included in the already generated contour line.
[0064]
Then, when the outline connection information DC is not yet stored in the memory 50, or as a result of referring to the outline connection information DC, any of the four outline sides of the target pixel is not included in the existing outline. If determined, the contour information generation unit 40a is instructed to generate a contour line composed of four contour sides of the target pixel. Thereby, the contour line information generation unit 40a generates a new contour line including four contour sides around the target pixel, and describes the connection state of the new contour line in the contour line connection information DC of the memory 50. The start edge of the new contour line is obtained and described in the contour management information DM.
[0065]
On the other hand, as a result of referring to the contour line connection information DC, when it is determined that any of the four contour sides of the target pixel is included in the existing contour line, the contour line information update unit 40b receives the contour line connection information DC and Command to update the contour management information DM.
[0066]
A process in which the contour line information update unit 40b updates the contour line connection information DC and the contour line management information DM will be described. FIG. 14 is a diagram illustrating a case where four contour sides of a target pixel are included in one existing contour line. In FIG. 14, the hatched pixel is the target pixel, the thin line is each contour side, and the bold line indicates the contour side included in the contour line. As shown in FIG. 14A, among the four contour sides a, b, c, and d of the target pixel, the contour side d is included in the existing contour line E1. When the target pixel is added in such a state, the existing outline E1 is updated as shown by the thick line in FIG.
[0067]
FIG. 15 is a diagram illustrating a process of updating the contour line E1 of FIG. 14A to the contour line E1 of FIG. 14B. 15A to 15E, the left diagram is the contour connection information DC stored in the memory 50, and the right diagram is a hypothesis that the contour is formed from the four contour sides of the target pixel. This is a virtual connection state of the contour side of the target pixel temporarily stored in the memory 50.
[0068]
First, when the target pixel shown in FIG. 14A is extracted, the outline connection information DC and the virtual connection information of the outline E1 are as shown in FIG. That is, the connection state of the contour line E1 is “e → d → f → g”. Further, the connection of the four contour sides of the target pixel is “a → b → c → d → a”.
[0069]
Then, by adding the pixel of interest this time, the contour side d is not included in the contour line E1. Therefore, the next contour side number for the contour side e having the contour side d as the next contour side (connection destination) in the contour line connection information DC is updated. That is, referring to the virtual connection state, the contour side d is connected to the contour side a, so that the next contour side number for the contour side e is rewritten to a. As a result, the outline connection information DC and the virtual connection information of the outline E1 are as shown in FIG.
[0070]
In addition, since the contour side d is not included in the contour line E1 by adding the pixel of interest this time, the next for the contour side c having the contour side d as the next contour side (connection destination) in the virtual connection information. Will update the contour side number. That is, referring to the contour line connection information DC, the contour side d is connected to the contour side f, so the next contour side number for the contour side c is rewritten to f. As a result, the outline connection information DC and the virtual connection information of the outline E1 are as shown in FIG.
[0071]
Since the contour side d is not included in the contour line E1 by adding the pixel of interest this time, the connection state for the contour side d is deleted from the contour side connection information DC. As a result, the outline connection information DC and the virtual connection information of the outline E1 are as shown in FIG.
[0072]
Then, the outline connection information DC and the virtual connection information of the outline E1 are represented by copying the connection states of the outline sides a, b, and c excluding the outline side d from the virtual connection information to the outline connection information DC. 15 (e). In the contour line connection information DC in FIG. 15E, the connection state of the contour line E1 as shown by the thick line in FIG. 14B is described. Accordingly, the virtual connection information may be deleted from the memory 50 when the processing so far is performed.
[0073]
Next, when the four contour sides of the target pixel are included in two or more pre-formed contour lines, the processing is almost the same as above, but there are also different portions. For example, FIG. 16 is a diagram illustrating a case where four contour sides of a target pixel are included in three existing contour lines. In FIG. 15, the hatched pixels are the target pixels, the thin lines are the respective contour sides, and the thick lines indicate the contour sides included in the contour lines. In FIG. 14A, the contour sides e, d, f, and g are included in the contour line E1, and the contour sides t, a, and s are included in the contour line E2, and the contour sides m, c, and n are included. Is included in the contour line E3. In the contour line E1, the contour sides are connected in the order of “e → d → f → g”, and in the contour line E2, the contour sides are connected in the order of “t → a → s”. In the contour line E3, the contour sides are connected in the order of “m → c → n”. When a pixel of interest indicated by diagonal lines is added in such a state, the contour lines E2 and E3 are integrated with the contour line E1 and disappear as shown by the thick lines in FIG.
[0074]
When the contour lines are integrated, the contour line information update unit 40b refers to the contour line management information DM of each of the contour lines E1, E2, and E3, and acquires the number of contour sides constituting each contour line (FIG. 4 ( b)). Then, only the contour line having the largest number of contour sides is left, and the other contour lines are eliminated. For example, if the contour line having the largest number of contour sides in FIG. 16A is E1, only the contour line E1 remains as shown in FIG. The lines E2 and E3 have disappeared.
[0075]
FIG. 17 shows contour connection information and the like in the state of FIG. 17A shows the outline connection information DC, and FIG. 17B is temporarily stored in the memory 50 when an outline is formed from the four outline sides of the pixel of interest. This is a virtual connection state of the contour side of the target pixel, and FIG. 17C shows the contour management information DM. In FIG. 17C, the column of the start side number and the start density value stores the number of the contour side and the density value of the target pixel when the respective contour lines E1 to E3 are generated. Here, these values are indicated by “−” for convenience.
[0076]
By adding the pixel of interest indicated by diagonal lines in FIG. 16A, the contour sides a, c, and d are not included in the contour line. Therefore, the next contour side number for the contour side t having the contour side a as the next contour side (connection destination) in the contour line connection information DC of FIG. 17A is updated. That is, referring to the virtual connection state in FIG. 17B, the contour side a is connected to the contour side b, so the next contour side number for the contour side t is rewritten to b.
[0077]
Next, since the number of contour sides of the contour E1 is the largest from FIG. 17C, the contours E2 and E3 disappear. Accordingly, all belonging contour numbers E2 and E3 in the contour connection information DC in FIG. 17A are rewritten to E1.
[0078]
Next, since the contour side a is not included in the contour line, the next contour for the contour side d having the contour side a as the next contour side (connection destination) in the virtual connection state of FIG. The edge number will be updated. That is, referring to the contour line connection information DC, the contour side a is connected to the contour side s, so the next contour side number for the contour side d is rewritten from a to s. Then, since the contour side d is not included in the contour line, the next contour side e with the contour side d as the next contour side (connection destination) in the contour line connection information DC in FIG. The contour side number will be updated. In other words, referring to the virtual connection state, since the contour side d is previously connected to the contour side s, the next contour side number for the contour side e is rewritten from d to s.
[0079]
Next, since the contour side c is not included in the contour line, the next for the contour side m having the contour side c as the next contour side (connection destination) in the contour line connection information DC in FIG. Will update the contour side number. That is, referring to the virtual connection state, the contour side c is connected to the contour side d, so the next contour side number for the contour side m is rewritten from c to d. However, since the contour side d is not included in the contour line, the contour side f connected to the contour side d in the contour line connection information DC is set as the next contour side number for the contour side m.
[0080]
Next, since the contour side c is not included in the contour line, the next contour side number for the contour side b having the contour side c as the next contour side in the virtual connection state of FIG. 17B is updated. Will be. That is, referring to the contour line connection information DC, the contour side c is connected to the contour side n, so the next contour side number for the contour side b in the virtual connection state is rewritten from c to n.
[0081]
Next, since the contour side d is not included in the contour line, the next for the contour side e having the contour side d as the next contour side (connection destination) in the contour line connection information DC in FIG. Is updated from d to s with reference to the virtual connection state. Further, since the contour side c is not included in the contour line, the next contour side number for the contour side m having the contour side c as the next contour side in the contour line connection information DC in FIG. Update from c to f. This process has already been performed once, but by performing the same process once again, the process of rewriting the connection state of the four contour sides a, b, c, d of the target pixel starts from which contour side. Even if it is done, the same result is obtained.
[0082]
Next, since the contour sides a, c, and d are not included in the contour line, the connection information for the contour sides a, c, and d is deleted in the contour line connection information DC in FIG. Then, the connection state for only the contour side b in the virtual connection state is written in the contour line connection information DC. In addition, since the contour lines E2 and E3 disappear due to this processing, a density value one higher than the density value of the target pixel is written in the end density value column for the contour lines E2 and E3 in the contour management information DM. The number of contour sides for these contour lines is set to 0, the effective density range for the contour lines E2 and E3 is determined, and end processing is performed for the contour lines E2 and E3. In addition, the number of contour sides for the contour E1 is updated in the contour management information DM. Note that when this process is performed, the virtual connection state of the target pixel is deleted from the memory 50.
[0083]
As a result of such processing, the outline connection information DC stored in the memory 50 is as shown in FIG. 18A, and the outline management information DM is as shown in FIG. 18B. That is, these pieces of information indicate a state in which the contour line E1 as shown in FIG. As a result of the processing as described above, the contour line E1 is updated to a contour line composed of 16 contour sides, and the contour lines E2 and E3 have been finished and disappeared.
[0084]
Next, a case where two or more contour sides of the four contour sides of the target pixel are included in the same contour line will be described. First, before performing the processing as described above, the contour line number to which the four contour sides of the target pixel belong is acquired, and the contours included in the same contour line among the four contour sides of the target pixel Get the edge number. Then, it is determined whether or not the connection states of the contour sides included in the same contour line among the contour sides of the target pixel are continuous. If the connection states of the contour sides included in the same contour line are continuous, the contour line is appropriately updated by performing the normal processing as described above. However, when the connection states of the contour sides included in the same contour line are not continuous, a hole is formed inside the contour line, and a new contour line is generated for the hole. .
[0085]
For example, when a pixel indicated by diagonal lines is extracted as a target pixel when a contour as shown in the left diagram of FIGS. 19A to 19D is formed, a contour as shown in the right diagram is generated. The In FIGS. 19A to 19D, circles indicate the start sides of the contour lines. As a result of performing the update processing of the contour line connection information DC as described above, when a hole-like contour line that newly forms a closed loop is generated inside the existing contour line, the start side with respect to the contour line Then, the start density value and the number of contour lines are obtained and registered in the contour management information DM. By performing such processing, it is possible to appropriately cope with the case where the contour line is separated and a new contour line is generated.
[0086]
Next, a case will be described in which there is a contour side set as the start side of the existing contour line among the four contour sides of the target pixel. First, when it is determined that there is a contour side set as the start side of the existing contour E1 among the four contour sides of the target pixel, an area for writing information of the new contour E2 in the contour management information DM is secured, The contents of the existing outline E1 are copied to the area. Then, a density value that is one higher than the density value of the pixel of interest with respect to the outline management information DM for the outline E2 is written to the end density value, and the number of outline sides is set to 0, thereby completing the end process for the outline E2. Apply. Then, the density value of the target pixel is overwritten in the column of the start density value of the outline management information DM for the existing contour E1, and the start side is extracted from the outline sides other than the four outline sides of the target pixel and overwritten. . By performing such processing, the contour line information can be satisfactorily updated even when the contour side set as the start side of the existing contour line is present among the four contour sides of the target pixel. .
[0087]
However, as shown in the left diagram of FIG. 20, when all of the four contour sides of the target pixel are included in the same pre-formed contour line, as the start side of the pre-made contour line among the four contour sides of the target pixel Even if there is a set contour side, it is not necessary to perform the processing as described above, and it is only necessary to perform the end processing for the existing contour line. Thereby, it updates to the appropriate outline as shown in the right figure of FIG.
[0088]
In the above description, the outline information update unit 40b conceptually shows the process of updating the outline connection information DC and the outline management information DM. Here, in order to make the above-described processing contents easier to understand, let us look at the processing process when the above processing is performed on the original image and the framed image shown in FIG. 21 to 30 are diagrams showing the contour management information DM and the contour connection information DC when the contour for each density value is extracted. In FIG. 21 to FIG. 30A, the number of each pixel indicates the density value for each pixel of the original image portion, and the density value is omitted for the density value 0 of the frame portion. In the following description, the contour side number shown in FIG. 7 is used to refer to the contour side of each pixel, and the coordinate value shown in FIG. 8 is used to refer to each pixel.
[0089]
First, the pixel-of-interest extraction unit 20 acquires the coordinate value 9 of the memory address 19 that is the pixel having the maximum density value from the buffer area sorted as shown in FIG. Referring to the coordinate values in FIG. 8, a pixel having a density value of 9 is extracted as a target pixel as shown in FIG. Since the four contour sides of the target pixel are not included in the existing contour lines as shown in FIG. 21A, a contour line composed of the four contour sides of the target pixel is generated, and the contour management information DM and the contour line are generated. The information is described in the connection information DC.
[0090]
In the contour management information DM, the contour line newly generated by the four contour sides of the pixel of interest is the contour line generated first, so the contour number is 1. Further, since the density value of the pixel having the coordinate value 2 (above the pixel of interest) among the adjacent pixels existing through the four contour sides of the pixel of interest is 0, the contour side above the pixel of interest is the start side. Since the contour side number of this contour side is 4, 4 is stored as the start side number. Further, since the density value of the target pixel at this time is 9, 9 is stored as the start density value, and 4 is stored as the number of contour sides.
[0091]
In the contour connection information DC, the connection state of the contour formed by the four contour sides of the target pixel is described, and 1 is stored in the associated contour number of each contour side.
[0092]
Next, the pixel-of-interest extraction unit 20 acquires the coordinate values 22 and 18 of the memory addresses 18 and 17 in the buffer area in FIG. 13B, sets these pixels as the pixel of interest, and forms an outline for the density value 8 As shown in FIG. As shown in FIG. 22A, since the contour sides of the two pixels having the density value 8 are not included in the contour line of the density value 9 that has already been generated, a new contour line is formed for each pixel. . The outline management information DM and the outline connection information DC are as shown in FIGS. 22B and 22C, respectively. Since the pixel having the coordinate value 22 is first processed as the target pixel, the contour line number of the contour line formed by the four contour sides of the pixel having the coordinate value 22 becomes 2, and the four pixels having the coordinate value 18 The outline number of the outline formed by the outline sides is 3.
[0093]
Next, the pixel-of-interest extraction unit 20 sequentially acquires the coordinate values stored in the memory addresses 16 to 13 in the buffer area in FIG. When the contour line is formed, it becomes as shown in FIG. As shown in FIG. 23A, the contour side above the pixel with the coordinate value 16 out of the four pixels with the density value 7 is included in the contour line (contour number 1) of the density value 9 already generated. Therefore, the outline connection information DC and the outline management information DM for the outline number 1 are updated. In addition, since each contour side for the other pixels is not included in the existing contour line, a new contour line is formed for each pixel. As a result, the contour management information DM and the contour connection information DC are as shown in FIGS. 23B and 23C, respectively.
[0094]
Next, the pixel-of-interest extraction unit 20 acquires the coordinate value 30 stored in the memory address 12 of the buffer area in FIG. 13B, makes this pixel the pixel of interest, and forms a contour line for the density value 6 When this is done, the result is as shown in FIG. As shown in FIG. 24 (a), the right contour side among the four contour sides of the pixel having the density value 6 is included in the contour line (contour number 5) of the density value 7 already generated. The outline connection information DC and the outline management information DM for the outline number 5 are updated. As a result, the contour management information DM and the contour connection information DC are as shown in FIGS. 24B and 24C, respectively.
[0095]
Next, the target pixel extraction unit 20 acquires the coordinate values 32 and 24 stored in the memory addresses 11 and 10 in the buffer area of FIG. When the contour line is formed, the result is as shown in FIG. As shown in FIG. 25A, the lower contour side of the pixel having the coordinate value 24 out of the two pixels having the density value 5 is included in the already generated contour line (contour line number 5). For the pixel having the coordinate value 32, the left contour side is included in the already generated contour (contour number 5), and the right contour side is already generated (contour number 4). ). Here, when the number of contour sides of contour numbers 4 and 5 is compared, contour number 4 is smaller, so that the contour line of contour number 4 is integrated with the contour line of contour number 5. As a result, the contour line management information DM for the contour line number 4 is subjected to end processing, the density value 6 is entered in the end density value column, and the number of contour sides is updated to zero. Further, the outline management information DM and the outline connection information DC for the outline No. 5 are updated. As a result, two pixels with density value 5 are included in the contour line of contour number 5, and contour management information DM and contour connection information DC are shown in FIGS. 25B and 25C, respectively. It becomes like this.
[0096]
Next, the pixel-of-interest extraction unit 20 acquires coordinate values 29, 17, and 11 stored in the memory addresses 9 to 7 of the buffer area in FIG. When the contour line is formed for the value 4, it is as shown in FIG. As shown in FIG. 26A, the right contour side of the pixel having the coordinate value 29 is included in the contour line of contour number 5, and the upper contour side is included in the contour line of contour number 2. Yes. Therefore, the contour line of contour number 2 is integrated with the contour line of contour number 5 and disappears. Further, the right contour side of the pixel having the coordinate value 17 is included in the contour line of contour number 3, the lower contour side is included in the contour line of contour number 5, and the left contour side is It is included in the contour line of contour number 1. Therefore, the contour lines of contour numbers 1 and 3 are integrated with the contour line of contour number 5 and disappear. Further, the right contour side of the pixel having the coordinate value 11 is included in the contour line of contour number 6, and the lower contour side is included in the contour line of contour number 5 generated by the previous integration. ing. Therefore, the contour line of contour number 6 is integrated with the contour line of contour number 5 and disappears. When three pixels having the density value 4 are processed as the target pixel in this way, the contour management information DM and the contour connection information DC are as shown in FIGS. 26B and 26C, respectively.
[0097]
Next, the target pixel extraction unit 20 acquires the coordinate values 25 and 15 stored in the memory addresses 6 and 5 of the buffer area in FIG. When the contour line is formed, the result is as shown in FIG. As shown in FIG. 27A, the upper, lower, and left contour sides of the pixel having the coordinate value 25 are included in the contour number 5. Accordingly, the contour line of contour number 5 is updated so that the pixel having the coordinate value 25 is included in the contour line of contour number 5. The right and lower contour sides of the pixel having the coordinate value 15 are included in the contour line of contour number 5, but the connection state is not continuous. For this reason, the process of updating the pixel having the coordinate value 15 to be included in the contour line of the contour line number 5 is a separation process, and a new contour line (contour line number 7) is placed inside the contour line of the contour line number 5. ) Is formed. When two pixels having the density value 3 are processed as the target pixel in this manner, the contour management information DM and the contour connection information DC are as shown in FIGS. 27B and 27C, respectively.
[0098]
Next, the target pixel extraction unit 20 acquires the coordinate values 23, 19, and 8 stored in the memory addresses 4 to 2 of the buffer area in FIG. When the contour line is formed for the value 2, it is as shown in FIG. As shown in FIG. 28A, all the contour sides of the pixel having the coordinate value 23 are included in the contour line of contour number 7 and the start side of the contour line is included. Therefore, an end process is performed on the contour line of contour number 7. In addition, since the upper and left contour sides of the pixel of the coordinate value 19 are included in the contour line of the contour line number 5, the contour line connection information DC and the contour line management information DM for the contour line number 5 are updated. The Further, since the right and lower contour sides of the pixel with the coordinate value 8 are included in the contour line of the contour number 5, the contour line connection information DC and the contour management information DM for the contour number 5 are updated. Is done. As a result, the contour management information DM and the contour connection information DC are as shown in FIGS. 28B and 28C, respectively.
[0099]
Next, the target pixel extraction unit 20 acquires the coordinate values 26 and 10 stored in the memory addresses 1 and 0 of the buffer area in FIG. When the contour line is formed, the result is as shown in FIG. As shown in FIG. 29A, since the contour side excluding the right side of the pixel having the coordinate value 26 is included in the contour line of contour number 5, the contour line connection information DC and contour line for contour number 5 The management information DM is updated. Further, since the contour side excluding the upper side of the pixel having the coordinate value 10 is included in the contour line of the contour number 5, the contour connection information DC and the contour management information DM for the contour number 5 are updated. . As a result, the outline management information DM and the outline connection information DC are as shown in FIGS. 29B and 29C, respectively.
[0100]
Finally, regarding the density value 0, when the density value 0 is used as a threshold value for contour extraction, the extracted contour line is the outer peripheral portion of the original image. Therefore, for the density value 0, it is only necessary to output a contour line connecting the outer peripheral portions of the original image without performing a process of extracting a pixel of interest and forming a contour line for each pixel. For the density value 0, end processing is performed on the contour management information DM for each contour line. The result of the termination process is shown in FIG.
[0101]
The process described above will be briefly described as follows. The target pixel is extracted pixel by pixel in descending order from the pixel indicating the maximum density value in the image. Then, when the four contour sides around the pixel of interest are not included in the existing contour line, the contour line including the four contour sides is set as a new contour line, and the contour indicating the connection state of the new contour line Generate line connection information. At this time, contour management information including the start side of the new contour line and the start density value having the density value of the target pixel as the start density of the effective density range of the new contour line is generated. On the other hand, when any of the four contour sides of the target pixel is included in the existing contour line, the contour line connection information and the contour line management information for the existing contour line are updated. Then, when at least one existing contour line that includes any contour side of the target pixel disappears due to the update, the density value of the target pixel with respect to the contour management information about the existing contour line that disappears Is stored as an end density value indicating the end density of the effective density range of the existing contour line.
[0102]
By performing such processing, when all pixels having the same density value are processed as the target pixel, the contour line of the density value is stored in the contour line connection information DC. Therefore, for example, when it is desired to output a contour line with respect to the density value 8, the desired density value 8 can be obtained by outputting the contour line connection information DC when all pixels indicating the density value 8 are processed as the target pixel. It is not necessary to process pixels having a density value of 7 or less. Further, when it is desired to output contour lines for all density values, if the contour line connection information DC is output every time processing of all pixels indicating the respective density values is performed, all the contours constituting the original image are output. The contour lines of all density values can be obtained by extracting and processing the pixels once.
[0103]
In this embodiment, when the contour line is formed for each density value, the contour line management information DM is generated / updated. Therefore, the processing up to the density value 1 is finally completed. In this case, even if the contour line connection information DC is deleted from the memory 50, if the contour line management information DM is stored, a contour line for an arbitrary density value can be extracted later.
[0104]
<5. Outline extraction>
Next, a process for extracting a contour line for each density value based on the contour management information DM after performing a process of forming a contour line while extracting the target pixel as described above one by one. explain. This process is performed in the contour extraction unit 60 of FIG.
[0105]
The contour extraction unit 60 refers to the contour line management information DM stored in the memory 50, finds the effective start side of the contour line at the density value to be extracted, and based on the predetermined rule (rule) from the start side. By tracing the contour, a contour line having a predetermined density value is extracted.
[0106]
Here, a rule for extracting a contour line will be described. FIG. 31 is a diagram illustrating a rule for specifying the traveling direction of the start side. The start side may be located along the main scanning direction X (ie, the horizontal direction) or may be located along the sub-scanning direction Y (ie, the vertical direction). When one start side is specified, the traveling direction of the start side is determined. For example, if the density values of two pixels existing across the start side are a and b, the start side is located in the vertical direction, and the density value a of the right pixel is greater than the density value b of the left pixel. If it is small, or if the start side is located in the horizontal direction and the density value a of the upper pixel is smaller than the density value b of the lower pixel, as shown in FIG. The traveling direction of the start side is set in the right direction. On the other hand, when the start side is located in the vertical direction and the density value a of the right pixel is larger than the density value b of the left pixel, or when the start side is located in the horizontal direction and the upper side When the density value a of the pixel is larger than the density value b of the lower pixel, as shown in FIG. 31B, the traveling direction of the start side is set downward or leftward.
[0107]
Next, FIG. 32 is a diagram showing a rule for specifying a traveling direction when tracing and forming a contour line. FIG. 32A shows an arrow whose traveling direction has already been determined. Here, if the density value to be extracted is a, the density value of the left pixel in front of the arrow is b, and the density value of the right pixel in front of the arrow is c, “a ≦ b and a ≦ c”. For example, if the path is in the left direction as shown in FIG. 32 (b) and “a> b and a ≦ c”, the path is in the straight direction as shown in FIG. 32 (c), and “a> c”. For example, as shown in FIG. When such a rule is set, a pixel larger than or equal to the density value to be extracted is always obtained on the right side with respect to the traveling direction at the time of contour extraction, and a pixel smaller than the density value to be extracted on the left side Is always obtained. Therefore, if the contour side is traced based on the above rule, a contour line having a desired density value can be extracted.
[0108]
An example in which a contour line having a density value of 3 is actually extracted based on the contour management information DM obtained for the original image in FIG. First, the effective density range for each contour line in the contour line management information DM of FIG. 30B obtained for the original image of FIG. 5 is as shown in FIG. In FIG. 33, contour numbers 5 and 7 include the density value 3 in the effective density range. Therefore, referring to the contour management information DM in FIG. 30B for the contour numbers 5 and 7, the start side numbers are 62 and 45, respectively. When the start side is specified from the contour side number assigned as shown in FIG. 7 and the contour side is traced from the start side based on the above rule, the result is as shown in FIG. FIG. 34A shows a contour line extracted from the start side of the contour side number 62, and FIG. 34B shows a contour line extracted from the start side of the contour side number 45. Then, by combining the two contour lines shown in FIGS. 34A and 34B, a contour line having a desired density value 3 can be obtained.
[0109]
In this way, the contour extraction unit 60 extracts the contour line from the start side and outputs the contour line. Therefore, when the contour extraction unit 60 outputs the contour line, it is not necessary to scan the image for each pixel, and the contour line can be output efficiently. That is, once the contour line management information DM is generated, by using it, the contour extraction process for an arbitrary density value can be performed efficiently, and the processing time can be shortened.
[0110]
<6. Processing sequence>
Next, a processing sequence in the case where the processing for generating / updating the above-described contour connection information DC and contour management information DM is performed by a computer or the like will be described. 35 to 38 are flowcharts showing a processing sequence for forming a contour line for each density value by a computer or the like.
[0111]
First, in step S10, preprocessing for the original image is performed. That is, in this pre-processing, the generation of the framed image as described above, the assignment of the contour side number, and the pixel sorting based on the density value are performed.
[0112]
Next, in step S11, the target pixel is extracted. The pixel of interest is extracted one pixel at a time in descending order of the density value from the pixel showing the maximum density value based on the sorting result in step S10.
[0113]
In step S12, it is determined whether or not the four contour sides of the target pixel extracted in step S11 are included in the existing contour line. This determination is performed by referring to the already generated outline connection information DC. If none of the four contour sides of the target pixel is included in the existing contour line, it is determined as “NO”, and the process proceeds to step S13. If it is included in the line, “YES” is determined, and the process proceeds to step S14.
[0114]
When the process proceeds to step S13, a new contour line composed of the four contour sides of the target pixel is generated, and contour connection information DC and contour management information DM are generated for the new contour line. And if the process of step S13 is performed, it will progress to step S41 mentioned later (connector A).
[0115]
When the process proceeds to step S14, the affiliated contour line number indicating the contour line including the contour side of the target pixel is acquired based on the contour line connection information DC. At this time, when two or more contour sides among the four contour sides of the pixel of interest are included in the existing contour line, the associated contour line numbers are acquired. And if the process of step S14 is performed, it will progress to step S15.
[0116]
In step S15, the belonging contour line number acquired in step S14 is verified, and it is determined whether or not there are contour sides included in the same contour line among the four contour sides of the target pixel. That is, when the belonging contour number is acquired for each of two or more contour sides included in the existing contour line among the four contour sides of the target pixel, there are contour sides indicating the same belonging contour number. In this case, “YES” is determined. In other cases, “NO” is determined. If "YES" is determined, the process proceeds to step S16, the connection state of the contour sides included in the existing contour line is stored in a memory or the like, and the process proceeds to step S17. If “NO” is determined, the process directly proceeds to step S17.
[0117]
In step S <b> 17, it is determined whether or not there is a contour side registered as the start side of the existing contour line among the four contour sides of the target pixel. This determination is performed by referring to the already generated contour management information DM and verifying whether the contour side numbers of the four contour sides of the target pixel are registered as the start side numbers of the existing contour lines. Is called. If there is a contour side registered as the start side of the existing contour line, “YES” is determined, and the process proceeds to step S18. If there is a contour side registered as the start side of the existing contour line, It is determined as “NO” and the process proceeds to step S21 (connector C).
[0118]
In step S18, it is determined whether or not the existing contour line including the contour side of the target pixel is composed of only the four contour sides of the target pixel. If "NO" is determined here, the process proceeds to step S19. If "YES" is determined, the process proceeds to step S20.
[0119]
When the start side is included in the contour side of the target pixel, it is necessary to update the start side of the contour line. This is because the start side included in the contour side of the target pixel becomes invalid by processing the target pixel. For this reason, when the start side is included in the contour side of the target pixel, a new contour line is generated, and the new contour line takes over the contour line connection information DC of the previous contour line. Since no start side is registered in the outline management information DM for the new outline, one outline side on the new outline (however, an outline side that is not in contact with the target pixel) is selected as the start side. Further, the start density value for the new contour line is set as the density value of the target pixel. Then, an end process is performed on the previous contour line. However, in practice, it is inefficient to change the information of the inherited new contour line. In this embodiment, the following processing is performed in step S19.
[0120]
An area for writing new outline information is secured in the outline management information DM, and the contents of the previous outline in which the invalid start side is registered in the area are copied (however, the outline numbers are different). . Then, with respect to the copied contour management information DM, a density value that is one higher than the density value of the target pixel is written in the end density value, and the number of contour sides is set to 0, whereby the end process is performed on the generated new contour line. Apply. On the other hand, the density value of the current target pixel is overwritten in the start density value field of the contour management information DM for the existing contour line in which the invalid start side is registered, and the four contours of the target pixel are used as the start side. Extract and overwrite from contour edges other than edges. By performing such a process, it is possible to set the newly valid starting edge by taking over the contents of the contour line in which the invalid starting edge is registered. As a result, the contour line information can be satisfactorily updated even when there is a contour side set as the start side of the existing contour line among the four contour sides of the target pixel. And if the process of step S19 is performed, it will progress to step S21 next (connector C).
[0121]
Further, when the start side is included in the contour side of the target pixel and the contour line is composed of only four contour sides of the target pixel, for example, as shown in FIG. is there. In this case, an end process may be performed on a contour line including only four contour sides of the target pixel. Accordingly, in step S20, an end process is performed on the contour line including the contour side of the target pixel. And if the process of step S20 is performed, it will progress to step S41 mentioned later (connector A).
[0122]
Next, in step S21, the contour management information DM is referred to, the contour having the largest number of contour sides among the contours including the contour side of the target pixel is obtained, and the contour number of the contour line is determined. get.
[0123]
In step S22, the affiliated outline number of the outline including the outline side of the target pixel in the outline connection information DC is rewritten to the outline number acquired in step S21. Next, in step S23, virtual connection states of the four contour sides of the target pixel are generated and held.
[0124]
In step S24, one of the four contour sides of the pixel of interest is included in the existing contour line, and in step S25, the contour side selected with reference to the contour connection information DC is selected as the next contour. The contour edge registered in the edge number is detected. In step S26, the next contour side which is the connection destination of the contour side selected in step S24 from the virtual connection state is detected. In step S27, it is determined whether or not the contour edge detected in step S26 is included in the existing contour line. If not included, the determination is “NO” and the process proceeds to step S28. In this case, “YES” is determined, and the process proceeds to step S29.
[0125]
In step S28, the contour side number detected in step S26 is stored in the contour side number next to the contour side detected in step S25. On the other hand, in step S29, the next contour side that is the connection destination of the contour side detected in step S26 is detected with reference to the contour line connection information DC, and this contour side is next to the contour side detected in step S25. The contour side of.
[0126]
In step S30, the contour connection information DC is referred to, and the contour side that is the connection destination of the contour side selected in step S24 is detected. In step S31, the virtual connection state is referred to, and the contour side registered as the next contour side is detected as the contour side selected in step S24.
[0127]
Then, in step S32, it is determined whether or not the contour side detected in step S31 is included in the existing contour side. If it is not included, “NO” is determined, and the process proceeds to step S33. In this case, “YES” is determined, and the process proceeds to step S34.
[0128]
In step S33, the contour side number detected in step S30 is stored in the contour side number next to the contour side detected in step S31. On the other hand, in step S34, the contour connection information DC is referred to, a contour side having the contour side detected in step S31 as the next contour side is detected, and the contour registered as the next contour side of the contour side is detected. Let the side be the contour side detected in step S30.
[0129]
The processing for updating the connection state of the contour sides for one contour side of the target pixel included in the existing contour line is completed in the processes of steps S24 to S34 described above. In step S35, it is determined whether or not there are other contour sides included in the existing contour line among the four contour sides of the target pixel. If “YES”, the same processing is again performed on the contour side. Return to step S24 to perform (connector E), and if “NO”, proceed to step S36 (connector F).
[0130]
In step S36, it is determined whether or not there is a connection state of the contour side stored in step S16. If there is a stored connection state, “YES” is determined and the process proceeds to step S37, and if there is no stored connection state, “NO” is determined and the process proceeds to step S41.
[0131]
In step S37, the stored connection state is read, and two contour sides included in the same contour line are extracted. In step S38, it is determined whether or not the connection state of these two contour sides is continuous. If not continuous, for example, as shown in FIG. 19, a hole-shaped contour line is formed inside the existing contour line, and the process proceeds to step S39. On the other hand, if it is continuous, a hole-shaped outline is not formed, and the process proceeds to step S40.
[0132]
In step S39, a start side is registered for a hole-shaped contour line formed inside the existing contour line, and set as a new contour line. Then, the process proceeds to step S40.
[0133]
In step S40, it is determined whether or not there are other combinations when extracting two contour sides included in the same contour line. If there are other combinations, a hole-shaped contour line is formed for those contour sides. In order to check whether or not it is formed, the process returns to step S37 again. If there is no other combination, “NO” is determined, and the process proceeds to step S41.
[0134]
The processing for the target pixel extracted by the processing in steps S11 to S40 is completed. In step S41, whether there is a pixel to be processed next as the target pixel is determined based on the memory address of the sorting result. If there is a next pixel, “YES” is returned to Step S11 (connector G). Further, when the processing for all the pixels is completed, “NO” is determined, and the process proceeds to step S42.
[0135]
In step S42, since there are no pixels of interest to be extracted, the end density value “0” is stored for all contour lines for which no end density value is stored in the contour management information DM at this time. That is, end processing is performed on all contour lines that have not been subjected to end processing. When this termination process is performed, the process proceeds to step S43.
[0136]
In step S43, the contour management information DM generated / updated by the above processing is saved, and the processing sequence ends.
[0137]
By performing the processing as described above, contour management information DM that can extract a contour for each density value is obtained. Further, when outputting a contour line for each density value in the process as described above, when the target pixel is extracted in step S11, the density value of the target pixel is examined to be different from the previous density value. For example, it may be configured to output the outline connection information DC at that time.
[0138]
As described above, if the contour extraction of an image is performed in the processing mode as described in this embodiment, the contour extraction processing for an arbitrary density value can be performed efficiently, and the processing time can be shortened. It is no longer obvious.
[0139]
<7. Modification>
In the above processing, the description has been given mainly on the processing for the 10-gradation image having the density value of 0 to 9, but the present invention is not limited to this, and the present invention can be applied to an image having an arbitrary gradation number Needless to say, it is applicable.
[0140]
Next, in the above description, in order to detect that a hole-like contour line is newly formed inside the existing contour line, it is detected whether or not the connection state of the contour sides continues. However, in addition to this, it is also possible to detect whether or not a hole-like contour line is newly formed by simply detecting an image pattern.
[0141]
Next, a general computer including a display unit, a keyboard, a CPU, a memory, etc., executes a predetermined program for each processing unit constituting the image contour extraction apparatus 100 in FIG. 2 described in the above embodiment. It may be realized by. In this case, the original image is stored in a magnetic disk or the like, and the contour line for each density value is executed by the CPU in the computer executing the image contour extraction program for performing each procedure shown in FIGS. Are sequentially formed, and the contour connection information DC and the contour management information DM are sequentially generated / updated in the memory. Then, when the computer performs the processing as described above, the time for the contour extraction processing by the computer is shortened. Note that the contour extraction program executed by the computer is recorded and stored in a portable recording medium such as a CD-ROM or a flexible disk, or a fixed recording medium such as a magnetic disk installed in the computer. The contour extraction program may be read from the recording medium and executed.
[0142]
【The invention's effect】
As described above, according to the first aspect of the present invention, the pixel of interest is extracted one by one in descending order from the pixel indicating the maximum density value in the image, and the contour line connection information, the contour line management information, In the process of updating, the contour line for each density value is formed. For this reason, when several contour lines for different density values are extracted, it is only necessary to perform the process of extracting all the pixels constituting the image as the target pixel and performing the process once. Therefore, the contour extraction process for an arbitrary density value can be performed efficiently, and the processing time can be shortened. Further, since the start side is included in the contour management information, the contour management information can be used when contour extraction of an arbitrary density value is performed later.
[0143]
According to the second aspect of the present invention, if pixels having the same density value in the image are processed as the target pixel, a contour line of the density value is obtained, and therefore the contour extraction process is efficiently performed. And the processing time can be shortened.
[0144]
According to the third aspect of the present invention, since a contour line having a predetermined density value can be extracted based on the stored contour line management information, contour line connection information is generated while extracting a target pixel again. There is no need. Accordingly, the contour extraction process can be performed efficiently, and the processing time can be shortened.
[0145]
According to the fourth aspect of the present invention, the pixels constituting the image can be efficiently rearranged in descending order or descending order of the density value. As a result, contour extraction processing can be performed efficiently, and processing time can be shortened.
[0146]
According to the fifth aspect of the present invention, the process of extracting the target pixel one pixel at a time in descending order from the pixel indicating the maximum density value in the image and updating the contour line connection information and the contour line management information. In FIG. 5, a contour line for each density value is formed. For this reason, when several contour lines for different density values are extracted, it is only necessary to perform the process of extracting all the pixels constituting the image as the target pixel and performing the process once. Therefore, the contour extraction process for an arbitrary density value can be performed efficiently, and the processing time can be shortened.
[0147]
According to the sixth aspect of the present invention, when the computer executes a predetermined image contour extraction program, the computer can efficiently extract the contour of the image, and the processing time can be shortened. .
[Brief description of the drawings]
FIG. 1 is a diagram showing the concept of outline generation / update in the present invention.
FIG. 2 is a block diagram showing a configuration of an image contour extracting apparatus as an embodiment of the present invention.
FIG. 3 is a diagram for explaining contour line connection information;
FIG. 4 is a diagram illustrating contour management information.
FIG. 5 is a diagram illustrating generation of a framed image.
FIG. 6 is a diagram illustrating assignment of contour side numbers.
FIG. 7 is a diagram showing outline side numbers for a 7 × 6 framed image.
FIG. 8 is a diagram illustrating coordinate values of each pixel in a 7 × 6 framed image.
FIG. 9 is a diagram illustrating a result of counting the number of pixels for each density value for an original image.
FIG. 10 is a diagram illustrating the cumulative number of pixels for each density value.
FIG. 11 is a diagram illustrating a sorting process based on density values.
FIG. 12 is a diagram illustrating a sorting process based on density values.
FIG. 13 is a diagram illustrating a cumulative pixel number and a buffer area for each density value subjected to sorting processing.
FIG. 14 is a diagram illustrating a case where four contour sides of a target pixel are included in one existing contour line.
FIG. 15 is a diagram showing a process of updating the contour line E1 in FIG. 14A to the contour line E1 in FIG. 14B;
FIG. 16 is a diagram illustrating a case where four contour sides of a target pixel are included in three existing contour lines;
FIG. 17 is a diagram showing contour line connection information and the like in the state of FIG.
FIG. 18 is a diagram illustrating an example of contour line connection information.
FIG. 19 is a diagram illustrating an example where a hole-shaped contour line is formed.
FIG. 20 is a diagram illustrating an example when a contour line disappears;
FIG. 21 is a diagram showing contour management information and contour connection information when contours for each density value are extracted.
FIG. 22 is a diagram showing contour management information and contour connection information when a contour is extracted for each density value.
FIG. 23 is a diagram showing contour management information and contour connection information when contours for each density value are extracted.
FIG. 24 is a diagram showing contour management information and contour connection information when contours for each density value are extracted.
FIG. 25 is a diagram showing contour management information and contour connection information when a contour is extracted for each density value.
FIG. 26 is a diagram showing contour management information and contour connection information when contours for each density value are extracted.
FIG. 27 is a diagram showing contour management information and contour connection information when contours are extracted for each density value.
FIG. 28 is a diagram showing contour management information and contour connection information when contours for each density value are extracted.
FIG. 29 is a diagram showing contour management information and contour connection information when contours for each density value are extracted.
FIG. 30 is a diagram showing contour management information and contour connection information when contours for each density value are extracted.
FIG. 31 is a diagram illustrating a rule for specifying a traveling direction of a start side.
FIG. 32 is a diagram illustrating a rule for specifying a traveling direction when tracing and forming an outline.
FIG. 33 is a diagram showing an effective density range for each contour line in the contour management information.
FIG. 34 is a diagram illustrating a process of tracing a contour side based on a rule from a start side.
FIG. 35 is a flowchart showing a processing sequence for forming a contour line for each density value by a computer or the like.
FIG. 36 is a flowchart showing a processing sequence for forming a contour line for each density value by a computer or the like.
FIG. 37 is a flowchart showing a processing sequence for forming a contour line for each density value by a computer or the like.
FIG. 38 is a flowchart showing a processing sequence for forming a contour line for each density value by a computer or the like.
[Explanation of symbols]
10 Preprocessing section
20 attention pixel extraction part
30 Contour edge determination unit
40 Contour line forming part
40a Outline information generation unit
40b Outline information update unit
50 memory
60 Contour extraction unit
70 storage unit
DM Outline management information
DC contour connection information

Claims (6)

A method for extracting the contour of an image,
(a) extracting a pixel of interest one pixel at a time in descending order from the pixel indicating the maximum density value in the image;
(b) When a plurality of contour sides around the pixel of interest are not included in the existing contour line, a contour line composed of the plurality of contour sides is set as a new contour line, and the connection state of the new contour line Is generated, and the start side of the new contour line selected from any one of the plurality of contour sides of the pixel of interest and the density value of the pixel of interest are used as the new contour line. Generating contour management information including a starting density value as a starting density of the effective density range of
(c) when any of the plurality of contour sides of the pixel of interest is included in a pre-formed contour, the step of updating the contour connection information about the pre-formed contour;
(d) When at least one pre-formed contour that includes any of the contour sides of the target pixel disappears due to the update, or when there is no target pixel to be extracted, Adding a density value of the target pixel to the contour management information as an end density value indicating an end density of an effective density range of the existing contour line;
A method for extracting an outline of an image, comprising: The image contour extracting method according to claim 1,
The contour line connection information generated / updated by repeatedly performing the steps (a) to (d), with the pixel having the same density value in the image as the target pixel, is output. Outline extraction method. The image contour extracting method according to claim 1,
(e) storing the outline management information generated by repeatedly performing the steps (a) to (d);
When extracting the contour line of the predetermined density value,
(f) identifying a contour line of the predetermined density value based on the effective density range indicated by the start density value and the end density value included in the contour line management information, and the contour for the contour line Obtaining the start side from line management information;
(g) sequentially extracting contour edges according to a predetermined rule from the start edge;
A method for extracting an outline of an image, comprising: The image contour extraction method according to any one of claims 1 to 3,
As a step performed prior to the steps (a) to (d),
(h) securing a recording area for recording pixel information for all the pixels constituting the image;
(i) searching for all the pixels constituting the image and counting the number of pixels for each density value;
(j) obtaining a cumulative number of pixels less than the density value for each density value based on the result of the counting;
(k) extracting pixels constituting the image one pixel at a time;
(l) Detecting the density value of the pixel extracted in the step (k), obtaining the cumulative pixel number corresponding to the density value, and at the position of the recording area corresponding to the cumulative pixel number value Recording the pixel information identifying the extracted pixels and updating the value of the cumulative pixel number;
Have
An image contour extracting method, wherein the steps (k) to (l) are repeatedly performed until all the pixels constituting the image are performed. An apparatus for extracting the contour of an image,
(a) pixel-of-interest extracting means for extracting a pixel of interest one pixel at a time in descending order from the pixel indicating the maximum density value in the image;
(b) When a plurality of contour sides around the target pixel are not included in the existing contour line, the contour line composed of the plurality of contour sides is set as a new contour line, and the connection state of the new contour line is indicated. The contour connection information is generated, and the start side of the new contour line selected from any one of the plurality of contour sides of the pixel of interest and the density value of the pixel of interest are used to determine whether the new contour line is valid. Contour information generating means for generating contour management information including a starting density value as a starting density of a density range;
(c) storage means for storing the contour line connection information and the contour line management information;
(d) When any one of the plurality of contour sides of the target pixel is included in the existing contour line, the contour line connection information stored in the storage unit is updated for the existing contour line In addition, when at least one existing contour line that includes any of the contour sides of the target pixel disappears due to the update, or when there is no target pixel to be extracted, Updating means for additionally updating the density value of the target pixel as an end density value indicating an end density of the effective density range of the existing contour line with respect to the outline management information stored in the storage means;
An image contour extraction apparatus comprising: A recording medium on which a program for extracting the contour of an image is recorded by a computer,
On the computer,
(a) a procedure for extracting a pixel of interest one pixel at a time in descending order from a pixel indicating the maximum density value in the image;
(b) When a plurality of contour sides around the pixel of interest are not included in the existing contour line, a contour line composed of the plurality of contour sides is set as a new contour line, and the connection state of the new contour line Is generated, and the start side of the new contour line selected from any one of the plurality of contour sides of the pixel of interest and the density value of the pixel of interest are used as the new contour line. A procedure for generating contour management information including a start density value as a start density of the effective density range of
(c) if any of the plurality of contour sides of the pixel of interest is included in a pre-formed contour, a procedure for updating the contour connection information for the pre-formed contour;
(d) When at least one pre-formed contour that includes any of the contour sides of the target pixel disappears due to the update, or when there is no target pixel to be extracted, A step of adding a density value of the target pixel to the contour management information as an end density value indicating an end density of an effective density range of the existing contour line;
The computer-readable recording medium which recorded the outline extraction program of the image for performing this.

Images