JPH06133160A - Picture area identification device - Google Patents

Abstract

PURPOSE:To provide the picture area identification device suitable for picture processing implementing preferable moire elimination in response to a type of picture data resulting from reading an original picture in which whether or not each area of the picture data is a dot area is accurately and automatically discriminated. CONSTITUTION:Picture data obtained by reading an original picture and converting it into a digital signal is read from an 8-line buffer memory (3) 301, a dot area discrimination section 205 divides the data into picture element blocks comprising 8X8 picture elements, applies DC transformation to each picture element block, each DCT coefficient value of obtained 8X8 DCT coefficient blocks is averaged as to a DCT coefficient block of several picture element blocks including a picture element block of discrimination object to obtain an evaluation block, a mean value of evaluation coefficients in plural 1st comparison areas along with a diagonal line of the evaluation block is obtained, and a mean value of evaluation coefficients in plural 2nd comparison areas along with a diagonal line of the evaluation block being a lower-degree order than the 1st comparison area is obtained, and the mean values of respective areas are compared, a switching signal to select the picture data smoothed by the smoothing processing section 201 or the picture data not smoothed is outputted to a signal changeover section 204.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to, for example, image characteristics of image data to be processed, particularly non-halftone images, when image data of a document read by a scanner is optimally processed according to the image characteristics. The present invention relates to an image area identification device capable of accurately determining a dot area and a halftone dot area.

[0002]

2. Description of the Related Art In an image forming apparatus such as a digital copying machine which reads an image of an original by using an image reading apparatus, converts it into a digital signal to obtain image data, and gives it to a recording apparatus to obtain a copied image. , Usually CCD in the image reading device
(Charge-coupled device) An image sensor is used to read an original image in a minute area, that is, for each pixel, and an analog electric signal obtained as an output of a CCD image sensor is A / D (analog / digital) converted,
After the digital signal obtained by A / D conversion is subjected to various kinds of processing, the processed signal is given to a recording device to obtain a copied image. By the way, in this type of image forming apparatus, an image is read in small pixel units by using a line sensor or the like as a CCD image sensor. Moire may occur on a recorded image due to interference with the array pitch (sampling period) of the line sensors.
As for this moire, there is known a technique of reducing the periodic component of the density by suppressing the generation of the moire by performing a moire removing process such as averaging of density data of a plurality of pixels.

[0003]

However, when such moire removing processing is performed on a character image or a halftone image, the image is blurred, so that a halftone image and a halftone image or a character image are mixed. It cannot be said that it is always an effective means for image data obtained by reading an existing document image. The present invention is intended to solve such a problem in the prior art, and is suitable for image processing for performing preferable moiré removal according to the image type of image data obtained by reading an original image, for example. It is an object of the present invention to provide an image area identification device that can accurately and accurately determine whether each area is a halftone dot area.

[0004]

According to the present invention, in order to solve the above-mentioned problems, image data corresponding to one screen is converted into N × N (N
Is an integer greater than or equal to 3), an image dividing unit that divides the pixel block into a plurality of pixel blocks, and two-dimensional orthogonal transformation is performed for each pixel block that is divided by the image dividing unit.
Transform coefficient calculation means for obtaining xN transform coefficients, and a transform coefficient block composed of N × N transform coefficients corresponding to the pixel block that is the subject of determination, with one pixel block as the subject of determination. The average value of the absolute values of the transform coefficients of the transform coefficient blocks corresponding to several pixel blocks in the vicinity of the pixel block to be determined is calculated for each transform coefficient, and the average value is used as the evaluation coefficient. An evaluation block calculation unit configured to form an evaluation block composed of evaluation coefficients, and an evaluation coefficient located on a diagonal line of an evaluation block from a low-order evaluation coefficient having the same vertical and horizontal orders in the evaluation block to a high-order evaluation coefficient. First evaluation area setting means for setting L (L is an integer of 2 or more) first evaluation areas configured by M (M is an integer of 2 or more) evaluation coefficients included therein;
For the evaluation area of, the L second evaluation areas configured by M ′ (M ′ is an integer of 2 or more) evaluation coefficients including evaluation coefficients in the lower order on the diagonal of the evaluation block are Second evaluation area setting means to be set, average value A [i] (i = 0 to L-1) of evaluation coefficients in the first evaluation area, and average value of evaluation coefficients in the second evaluation area Mean value calculating means in the evaluation area for obtaining B [i], mean value A [i] and mean value B [i]
To determine whether or not there is a combination of A [i]> B [i], and a determination signal indicating whether or not the pixel block to be determined according to the determination result is a halftone dot region. Is provided with a halftone dot area determination means.

[0005]

The image dividing means divides the image data corresponding to one screen into a plurality of pixel blocks composed of N × N pixels. The transform coefficient calculating means performs a two-dimensional orthogonal transform for each pixel block divided by the image dividing means to obtain N × N transform coefficients. The evaluation block calculation means sets one pixel block as a determination target, and a transform coefficient block composed of N × N transform coefficients corresponding to the determined pixel block, and the vicinity of the determined pixel block. The average value of the absolute values of the transform coefficients of the transform coefficient blocks corresponding to the several pixel blocks is calculated for each transform coefficient, and the obtained average value is used as an evaluation coefficient, and an evaluation block composed of N × N evaluation coefficients is obtained. Let me configure. The first evaluation area setting means is composed of M evaluation coefficients including the evaluation coefficients located on the diagonal of the evaluation block from the low-order evaluation coefficient having the same vertical and horizontal orders in the evaluation block to the high-order evaluation coefficient. L
First individual evaluation areas are set. The second evaluation area setting means sets, for each first evaluation area, L second evaluations configured by M ′ evaluation coefficients including evaluation coefficients in a lower order direction on a diagonal line of the evaluation block. Set the area. The evaluation area average value calculating means calculates the average value A [i] of the evaluation coefficients in the first evaluation area and the average value B of the evaluation coefficients in the second evaluation area.
Find [i]. The halftone dot area determination means compares the average value A [i] and the average value B [i] to determine whether there is a combination of A [i]> B [i], and determines according to the determination result. A determination signal indicating whether the target pixel block is a halftone dot area is output.

[0006]

Embodiments of the present invention will now be described in detail with reference to the drawings. According to the present invention, image data is orthogonally transformed by using a cosine function for each 8 × 8 block, that is, DCT (Discret)
e Cosine Transform) is performed to detect the frequency characteristic of the halftone dot image to determine whether the image data is in the halftone dot area. FIG. 1 is a block diagram showing the outline of the flow of images (signals) in the digital copying machine according to the first embodiment of the present invention. First, in the image reading unit (101), the original image is read by the CCD image sensor, the reflected light of the original image is converted into an analog electric signal, and A /
The D converter (102) converts an analog electric signal into a digital signal. The image processing unit (1
In 03), the halftone dot area is subjected to smoothing processing, and the area other than the halftone dot is transmitted to the image recording unit (104) without being subjected to smoothing processing. FIG. 2 is a block diagram of the internal circuit of the image processing unit (103). The image data signal output from the A / D conversion unit (102) is subjected to smoothing processing in the smoothing processing unit (201), or is temporarily subjected to 8 times without being smoothed.
The 8-line buffer memories [1] (202) and [2] (203) that store the image data for one line are stored, and the image data for each line is read out from them and is then read by the signal switching unit (204). Is entered. Halftone dot determination unit (2
If the image data input in 05) is determined to be the image data in the halftone dot area, the signal switching unit (204) reads the smoothed image from the 8-line buffer memory [1] (202). If the signal is judged to be image data other than the halftone dot area, 8 line buffer memory [2]
A switching signal for switching each pixel is input to the signal switching unit (204) so as to output the unprocessed image signal read from (203).

FIG. 3 is a block diagram of the internal circuit of the halftone dot area judging unit (205). The image data input from the A / D converter (102) is first stored in the 8-line buffer memory [3] (301) for 8 lines. The image data read from the 8-line buffer memory [3] (301) is stored in the DCT conversion unit (302) as shown in FIG.
It is divided into pixel blocks of × 8 pixels (a) and (b), and DCT conversion is performed for each pixel block. The DCT transform operation is performed for each pixel block by obtaining the DCT transform coefficient for the image data according to the following equation. Y _uv = 2C (u) C (v) / N ・ Σ _{i = 0} ⁷ Σ _{j = 0} ⁷ X _ij COS [(2i + 1) uπ / 2N] ・ COS [(2j + 1) vπ / 2N] , N = 8: u, v = 0 to 7: C (w) = 2 ^-1/2 ; w = 0 = 1; w = 1 to 7 (1) In the formula (1), X _ij is a pixel The luminance value of each pixel of the block, _Yuv is the value of the transformed DCT coefficient. (0,
The DCT coefficient Y _uv of the index of 0) is called a direct current (DC) component, and represents the average luminance in the pixel block, and is (0,0).
Coefficients other than are called alternating current (AC) components, and represent the magnitude of the spatial frequency component of the luminance of the pixels included in the pixel block. As can be seen from the equation (1), the DCT coefficient Y _uv represents a higher spatial frequency component as the value of (u, v) increases. D output from the DCT conversion unit (302)
The CT coefficient data is stored in the 8-line buffer memory [4] (30
It is input to either 3) or [5] (304).
These two 8-line buffer memories [4] (303),
[5] In (304), DC is replaced every 8 lines of processing.
The T coefficient Y _uv is accumulated. When all the DCT coefficients Y _uv for 8 lines are accumulated in the 8 line buffer memory [4] (303) or [5] (304), the average value processing unit (305) outputs 8 × 8 DCT coefficients for every several blocks. The DCT coefficient data of the DCT coefficient block composed of the block is read, and several blocks of the DCT coefficient block including the determination target pixel block are extracted. As will be described later, the average value of the absolute values of the components of the extracted DCT coefficient blocks of several blocks is calculated, and the average value is used as an evaluation block to determine whether or not the pixel block to be determined is a halftone dot region.

FIG. 5 is an explanatory diagram showing how to obtain an evaluation block in the average value processing unit (305). As shown in the figure, the five coefficient blocks before and after including the determination target pixel block and the five coefficient blocks of the line immediately before are extracted, and the average value of the absolute values of the respective DCT coefficients Y _uv included in these extracted blocks is calculated. Ask. The evaluation coefficient of this evaluation block is Y ′ _uv , and the DCT coefficient value of each block of the 10 coefficient blocks shown in FIG. 5 is Y [i] _uv (i is the coefficient block number shown in FIG. 5).
Then, the evaluation coefficient Y ′ _uv component is obtained by the following equation. Y ′ _uv = (Σ _{i = 0} ⁹ _│Y [i] _uv │) / 10 (2) The determination unit (306) uses the evaluation coefficient Y ′ _uv to determine the pixel block to be determined as a halftone dot region. Or not.
The determination as to whether this is a halftone dot region is performed by detecting the characteristic spatial frequency of the halftone dot image. In the halftone image, the dots are evenly arranged in the vertical and horizontal directions in the minute area, so that the DCT coefficient Y _uv component obtained by the DCT transform has a large component in which the vertical and horizontal spatial waves are added. Therefore, the value of the evaluation coefficient Y ′ _uv component obtained by DCT transforming the halftone image is the evaluation block shown by the 8 × 8 grid in FIG. There will be a peak on the diagonal line indicated by.
FIG. 6B is a specific example showing a three-dimensional graph of the evaluation coefficient Y ′ _uv component when the image data of the halftone image is DCT-transformed. As in this example, the image data of the halftone image is set to DC.
The coefficient Y ′ _uv component of the evaluation block when T-transformed generally has a peak on the diagonal line of the evaluation block. In this way, in order to detect the evaluation coefficient Y ′ _uv component characteristic of the halftone image, the determination unit (306) compares the values of the regions of the evaluation coefficient Y ′ _uv component in the evaluation block by the method described below. To do.

FIG. 7 shows a comparison area of the evaluation blocks to be compared by the judging section (306). FIG. 7A
And (b) represent two comparison areas in the same evaluation block. In FIG. 7A, the average value of the evaluation coefficient Y ′ _uv of the area composed of the evaluation block indexes (2, 2), (3, 2), (2, 3), and (3, 3) is A.

[0], indices (3, 3), (4, 3), (3, 4), and (4, 4), the average value of the evaluation coefficient _Y'uv is A [1],
Indicators (4,4), (5,4), (4,5), (5,5)
The average value of the evaluation coefficient Y ′ _uv of the area
[2], indices (5, 5), (6, 5), (5, 6),
The average value of the evaluation coefficient Y ′ _uv of the area constituted by (6, 6) is A [3], the indexes (6, 6), (7, 6), (6
7), the average value of the evaluation coefficient Y ′ _uv of the area composed of (7, 7) is represented as A [4]. In addition, in FIG.
Evaluation block indicators (1,1), (2,1), (3
1), (1, 2), the average value of the evaluation coefficient Y ′ _uv of the area composed of (1, 3) is B

[0], index (2, 2),
The average value of the evaluation coefficient Y ′ _uv of the area composed of (3,2), (4,2), (2,3) and (2,4) is B [1], the index (3,3), (4,3), (5,3), (3,4),
The average value of the evaluation coefficient Y ′ _uv of the region composed of (3, 5) is B [2], the indices (4, 4), (5, 4), (6
4), (4,5), the average value of the evaluation coefficient Y ′ _uv of the region composed of (4, 6) is B [3], the index (5, 5),
The average value of the evaluation coefficient Y ′ _uv of the area composed of (6, 5), (7, 5), (5, 6) and (5, 7) is represented by B [4]. Determination unit (306) 'by comparing the average value of _uv, evaluation coefficient Y on the diagonal of the evaluation block' evaluation coefficient Y of these comparisons in the region for detecting the peak of _uv. That is, A
When there is a combination of [i]> B [i], it is assumed that a peak of the evaluation coefficient Y ′ _uv exists on the diagonal line of the evaluation block, and the determination target pixel block is determined to be a halftone dot area.

Although the halftone dot area determination may be performed in this way, if such a halftone dot area determination is performed, a pixel in which each value of the evaluation coefficient _Y'uv of the original image with a flat density change is very small. Due to small noise in the image portion of a block, A [i]> B [i] may occasionally occur even though it is not a halftone dot region. In such a case, an erroneous determination of the region is caused. Therefore, a predetermined threshold value th1 is previously obtained by an experiment, and A
When [i]> B [i] and A [i]> th1, the determination target pixel block is determined as a halftone dot region, so that the precision of the halftone dot region determination is further improved.

The judgment section (306) judges whether or not the pixel block to be judged is a halftone dot area by the above-mentioned halftone dot area judgment processing, and if it is a halftone dot area, an 8-line buffer memory [6] (307). ), 1 is written as the dot area determination signal of the determination target pixel block, and 0 is written in other than the dot area.
Write. In this way, if the halftone dot area determination signals for eight lines are written in the eight-line buffer memory [6] (307), the eight-line buffer memory [1] (20
2), [2] This halftone dot area determination signal is read out in synchronization with the image data stored in (203) and output to the signal switching unit (204). As a result, when the halftone dot area judgment signal 1 is output, the judgment target pixel block is the halftone dot area, so the image read from the buffer memory [1] (202) in which the smoothed signal is stored. Since the data is other than the halftone dot area when the halftone dot area determination signal 0 is output, the image data read from the buffer memory [2] (203) in which the image data that has not been smoothed is stored is The signal switching unit (204) selects and outputs. In this way, since the image processing unit (103) smoothes only the halftone dot area, the moire is removed from the recorded image in the halftone dot area, and the moire removal processing is performed on the recorded image in the image area other than the halftone dot. Since there is no influence, a good recorded image can be obtained.

By the way, when the above-described halftone dot area determination processing is performed, in the halftone dot area determination performed by the determination unit (306), an erroneous determination sometimes occurs at the edge portion of the character image. In order to prevent such erroneous determination, the erroneous determination can be removed to perform more accurate halftone dot area determination. FIG. 8 is a block diagram of the internal circuit of the halftone dot area determination unit (205) according to the second embodiment of the present invention which enables elimination of erroneous determination. In addition, the same reference numerals are given to the same or similar parts as those of the first embodiment, and the description thereof will be omitted. The operation up to the determination unit (306) is exactly the same as in the first embodiment. When the determination unit (306) determines that the pixel block to be determined is a halftone dot region, a halftone dot region determination signal of 1 and 0 when it is determined to be other than the halftone dot region is a determination result buffer memory [1] (308). Or [2] (309). Halftone dot area determination signals are alternately stored in the two determination result buffer memories [1] (308) and [2] (309) every processing of eight lines. When all the halftone dot area determination signals for eight lines are stored, the determination result buffer memory [1] (308), [2] (309)
The halftone dot area determination signal stored in is read out, input to the erroneous determination removing section (310), and evaluated.

The erroneous decision removing process in the erroneous decision removing section (310) will be described. FIG. 9 shows a reference block group referred to in the erroneous determination removal processing. As shown in FIG. 9, the erroneous judgment evaluation results of the reference block groups C and D including the nearest pixel block to the judgment target pixel block are C [i] and D [i] (i = 0 to 5), When the sum of these erroneous judgment evaluation results is S, the sum S of erroneous judgment evaluation results
Is smaller than the predetermined threshold value th2 obtained by the experiment, the sum of the erroneous judgment evaluation results C [i] is not 6, and the sum of the erroneous judgment evaluation results D [i] is not 6, that is, S = Σ _{i = 0} ⁵ C [i] + Σ _{i = 0} ⁵ D [i] S <th2: Σ _{i = 0} ⁵ C [i] ≠ 6: Σ _{i = 0} ⁵ D [i] ≠ 6 (3) network It is determined to be outside the point region, the determination signal of the determination target pixel block is set to 0, and in other cases, the determination signal is not rewritten. The above-mentioned erroneous judgment removal processing utilizes the property of the image that halftone dot areas do not exist in isolation, and the judgment result of the area erroneously judged as halftone dot areas other than the halftone dot areas such as characters. Is to be removed. If the determination result of the determination target block is 1 in the erroneous determination removal processing of the erroneous determination removal unit (310) and it is determined that the block is a halftone dot region, the value of the halftone dot region determination signal of the pixels for one block is set to 1. , Those values are written in the 8-line buffer [6] (307), otherwise, the value of the dot area determination signal of the pixel of the determination target pixel block is set to 0, and those values are set in the 8-line buffer [6]. ] (307) Write. When the halftone dot area determination signals for eight lines are written in the eight line buffer memory [6] (307), the halftone dot area determination signals are transferred to the eight line buffer memories [1] (202), [2] (203). The signal is output to the signal switching unit (204) in synchronization with the pixel data stored in the. As a result, when the halftone dot area determination signal 1 is output, since it is a halftone dot area, the buffer memory [1] (20) in which the smoothed signal is stored.
Since the pixel data of 2) is outside the halftone area when the halftone area determination signal is 0, the pixel data of the buffer memory [2] (203) in which the unprocessed pixel data is stored is the signal switching unit ( 204) and output. In this way, since the image processing unit (103) performs smoothing only on the halftone dot area, moire in the halftone dot area of the recorded image is removed. On the other hand, in other image areas, a good recorded image is output without being affected by the moire removal processing.

[0014]

As described above, according to the first aspect of the invention, the image data corresponding to one screen is divided into a plurality of pixel blocks composed of N × N pixels, and each pixel block is divided. 2D orthogonal transformation is performed to obtain N × N transform coefficients, and one pixel block is determined as a determination target, and a corresponding transform coefficient block configured by N × N transform coefficients is determined as the determination target. Evaluation coefficient composed of N × N evaluation coefficients obtained for each conversion coefficient by averaging absolute values of the conversion coefficients of the conversion coefficient blocks corresponding to several pixel blocks in the vicinity of the pixel block. L blocks that are configured to include M evaluation coefficients including evaluation coefficients located on diagonal lines of the evaluation block from the low-order evaluation coefficient having the same vertical and horizontal orders to the high-order evaluation coefficient in the evaluation block. First evaluation area Set, for each first evaluation area, L second evaluation areas composed of M ′ evaluation coefficients including evaluation coefficients in the lower order on the diagonal of the evaluation block, The average value A [i] of the evaluation coefficients in the first evaluation area and the average value B of the evaluation coefficients in the second evaluation area
[I] is calculated, the average value A [i] is compared with the average value B [i], it is determined whether or not there is a combination of A [i]> B [i], and the determination result is determined according to the determination result. Since the judgment signal indicating whether the target pixel block is a halftone dot area is output, the image data of the halftone dot area is selected from the image data of the image in which halftone dots, photographs and characters are mixed. Since it is possible to accurately discriminate only the area, it can be suitably applied to an image processing apparatus that performs image processing by distinguishing a halftone dot area from another area. According to the second aspect of the present invention, the determination signal indicating whether or not the pixel block to be determined is the halftone dot area is obtained by comparing the average value A [i] with the average value B [i], Whether a combination of [i]> B [i] exists,
Further, since it is determined whether the average value A [i] is larger than a predetermined threshold value and the result is output according to the determination result, the image data of the image other than the halftone dot area is determined to be the halftone dot area. The number of erroneous determinations is reduced, and the accuracy of the halftone dot region determination can be further improved. According to the invention described in claim 3, a determination signal of 1 is output when the pixel block that is the determination target is determined to be a halftone dot area, and a determination signal of 0 is output when it is determined to be a non-halftone dot area, Except for the pixel blocks adjacent to the pixel block that is the determination target, K reference pixel blocks that are close to the left and right sides of the pixel block that is the determination target are extracted, and the reference pixel blocks on the left side are extracted. The output signal of the halftone dot area determination is C [i] (i =
0-K-1), and the output signal of the dot area determination for the reference pixel block on the right side is D [i] (i = 0-K-1), the output signals C [i] and D [i] When the total sum is smaller than the predetermined threshold value and the total sums of the output signals C [i] and D [i] are not K, the pixel block to be judged is judged to be a halftone dot area, and the judgment signal of 1 is outputted. Signal 0
When the output signals C [i] and D [i] do not satisfy the above condition, the determination signal is not changed. Therefore, the erroneous determination that a portion other than the halftone dot area is the halftone dot area is made. It is further reduced, and the accuracy of the halftone dot area determination can be further improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing an outline of a flow of image signals in a digital copying machine according to a first embodiment of the present invention.

FIG. 2 is a block diagram of an internal circuit of an image processing unit.

FIG. 3 is a block diagram of an internal circuit of a halftone dot area determination unit.

FIG. 4 is an explanatory diagram showing a state in which image data is divided into pixel blocks of 8 × 8 pixels and DCT conversion is performed for each pixel block.

FIG. 5 is an explanatory diagram showing how to obtain an evaluation block in an average value processing unit.

FIG. 6 is an explanatory diagram showing a diagonal line of an evaluation block and a three-dimensional graph showing a specific example of a coefficient Y ′ _uv component of the evaluation block.

FIG. 7 is an explanatory diagram showing a comparison area of evaluation blocks.

FIG. 8 is a block diagram of an internal circuit of a halftone dot area determination unit according to a second embodiment of the present invention.

FIG. 9 is an explanatory diagram showing a reference block group referred to in the erroneous determination removal processing.

[Explanation of symbols]

Reference Signs List 101 image reading unit 102 A / D conversion unit 103 image processing unit 104 image recording unit 201 smoothing processing unit 202, 203, 303, 304, 307 8 line buffer memory 204 signal switching unit 205 halftone dot area determination unit 302 DCT conversion unit 305 Average value processing unit 306 Determination unit 310 False determination removal unit

Claims (3)

[Claims] 1. Image data corresponding to one screen is N × N.
(N is an integer of 3 or more) An image dividing unit that divides the pixel block into a plurality of pixel blocks, and two-dimensional orthogonal transformation is performed for each of the pixel blocks divided by the image dividing unit, and N × A transform coefficient computing means for finding N transform coefficients, and a transform coefficient block composed of N × N transform coefficients corresponding to the pixel block which is the subject of decision, with one pixel block as the subject of decision , An average value of the absolute values of the transform coefficients of the transform coefficient blocks corresponding to the several pixel blocks in the vicinity of the pixel block that is the determination target is obtained for each transform coefficient, and this is used as an evaluation coefficient, and N Evaluation block calculation means for forming an evaluation block composed of × N evaluation coefficients, and before reaching from a low-order evaluation coefficient having the same vertical and horizontal orders in the evaluation block to a high-order evaluation coefficient A first (L is an integer of 2 or more) first evaluation area including M (M is an integer of 2 or more) evaluation coefficients including evaluation coefficients located on the diagonal of the evaluation block is set. The evaluation area setting means and M ′ (M ′ is an integer of 2 or more) evaluation coefficients including evaluation coefficients in a lower order direction on a diagonal line of the evaluation block for each of the first evaluation areas. Second evaluation area setting means for setting L second evaluation areas, and an average value A [i] (i = 0 to L-1) of evaluation coefficients in the first evaluation area, The evaluation area average value calculating means for obtaining the average value B [i] of the evaluation coefficients in the second evaluation area is compared with the average value A [i] and the average value B [i], and A [ i]> B [i] combination is determined, and the pixel block determined as a determination target according to the determination result is a halftone dot. An image area identification device comprising halftone dot area determination means for outputting a determination signal indicating whether or not it is a dot area. 2. The halftone dot area determination means compares the average value A [i] with the average value B [i] to determine whether there is a combination of A [i]> B [i], and further It is determined whether or not the average value A [i] is larger than a predetermined threshold value, and a determination signal indicating whether or not the pixel block to be determined is a halftone dot area is output according to the determination result. Claim 1
The image area identification device described. 3. The halftone dot area judging means outputs a judgment signal of 1 when the pixel block to be judged is a halftone dot area and outputs a judgment signal of 0 when it is judged to be a non-halftone dot area, and the judgment is made. Except for the pixel block adjacent to the target pixel block, K (K is an integer of 2 or more) reference pixel blocks that are adjacent to the left side and the right side of the pixel block to be determined, respectively. C [i] (i = 0 to K−1) for the output signal of the halftone dot area determination unit for the left reference pixel block, and the output of the halftone dot area determination unit for the right reference pixel block. Signal D
[I] (i = 0 to K-1), the output signal C [i]
And the sum of D [i] is smaller than a predetermined threshold, the output signal C
If the sums of [i] and D [i] are not K, the one dot determination signal output by the halftone dot region judging means judging the halftone dot region for the pixel block to be judged is output. 0 and output signals C [i] and D
3. The image area identifying device according to claim 1, further comprising an erroneous judgment removing means that does not change the judgment signal output by the halftone dot area judging means when [i] does not satisfy the above condition.