JP3216165B2 - Image region identification device and image region identification method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image area identifying apparatus and an image area identifying method for identifying different types of image areas such as a binary image area and a halftone area.

[0002]

2. Description of the Related Art In a facsimile or a digital copy, image information on an original is read using an image sensor, quantized according to the density of the image, and reproduced using a recording device or a display device. Is going.

In such an apparatus, when a binary image area such as a character or a line drawing and a halftone area such as a halftone picture are mixed in a document, a simple binarizing process is performed on the binary image area. Further, dither processing or error diffusion processing is performed on the halftone area.

[0004] A typical prior art device for distinguishing between a binary image region and a halftone region utilizes the density difference of pixels in a block. This device divides an image to be processed into a plurality of blocks each including a plurality of pixels, obtains a pixel having a maximum density level and a pixel having a minimum density level in each block, and determines a difference between these density levels and a predetermined density level. The threshold value is compared, and if it is larger than the threshold value, it is determined to be a binary image area, and otherwise, it is determined to be a halftone area.

[0005]

However, in the above-described conventional area discriminating apparatus, when discriminating an original having a poor paper quality, such as a newspaper, for example, the background density of the original is large, so that the density difference in the character portion becomes small, and There is a risk that the area may not be determined as a character area.

The present invention has been made in order to solve such a conventional problem, and to which of a plurality of different image areas a processing target image area belongs, for example, a binary image area and a halftone area. It is an object of the present invention to provide an image area identification device and an image area identification method that can identify with high accuracy.

[0007]

Means for Solving the Problems] image region identification apparatus according to claim 1, detecting means for detecting a feature amount from the concentration level of at least one line of pixels of the document that contains the processed image, has been detected Original text and halftone dots
Determine whether the image is a mixed image of photos or a mixed image of text and photos
Document mode determining means, and
It is characterized in that region identification is performed using the output .

[0008]

[0009]

[0010]

[0011]

According to a second aspect of the present invention , there is provided an image area identifying method , comprising: a detecting step of detecting a characteristic amount from a density level of pixels of at least one line of a document including an image to be processed ;
The original is a mixed image of characters and halftone photos based on the output features.
Original mode for judging whether images or mixed images of text and photos
And a judgment step.
It is characterized in that region identification is performed using force .

[0013]

[0014]

[0015]

[0016]

[0017]

[0018]

[0019]

[0020] In the image area identification method according to claim 1 of the image region identification apparatus and claim 2 of the structure arrangement, the document is, after it is determined whether a character and halftone photograph mixed image or a character and a photograph of the mixed image Since region identification is performed, identification can be performed more accurately.

[0021]

[0022]

[0023]

[0024]

[0025]

[0026]

[0027]

[0028]

[0029]

FIG. 1 is a block diagram showing an embodiment of an image area identifying apparatus according to the present invention. An image of the first predetermined number of lines of the document is read by an image sensor (not shown),
It is quantized by an A / D converter (not shown) and input to the decoder 1. Counters C1 to Cn provided at the next stage of the decoder 1 correspond to preset n (n is a positive integer) density levels. The decoder 1 selects one of the counters C1 to Cn according to the density level of the input image data, and increases the value of the selected counter by one. Accordingly, the count values of the counters C1 to Cn indicate the distribution of the density levels of the pixels of the first predetermined number of lines of the document, that is, the frequency of each density level of the histogram. FIG. 2 shows an example of the histogram thus detected.

The background density detecting circuit 3 outputs a peak value which first appears in the low-frequency density of the histogram indicated by the outputs of the counters C1 to Cn as the background density of the document. The background density output from the detection circuit 3 is supplied to a threshold selector 4. The threshold selector 4 reads a threshold used for area identification from the threshold memory 5 according to the input background density, and outputs it to the area identification circuit 6.

The original mode discriminating circuit 7 determines the pattern based on the distribution state of the histogram output from the counters C1 to Cn.
It is determined whether the original is a mixed image of characters and halftone photographs or a mixed image of characters and photographs by matching, and the determined original mode is output to the area identification circuit 6. AND gate 8
Receives a high-level signal output when the document mode determination circuit 7 completes the mode determination and a high-level signal output when the threshold selector 4 outputs a threshold, and outputs a high-level signal. The switch 9 is turned on when receiving a high level output from the AND gate 8. The area identification circuit 6 extracts a characteristic amount from image data of lines following the first predetermined number of lines of the original input through the switch 9, and outputs an original mode and a threshold value output from the determination circuit 7. Based on the threshold value read out by the selector 4, the processing target image identifies a force belonging to either a binary image area or a halftone area, and outputs an identification result.

FIG. 3 is a block diagram showing the configuration of another embodiment of the image area identification apparatus of the present invention. The image of the first predetermined number of lines of the document is read by an image sensor (not shown), quantized by an A / D converter (not shown), and input to comparators 11 and 12. Comparator 1
1 compares the density level of the input image data with the maximum density level stored in the maximum density memory 13 and filters out the higher density level. Maximum density memory 13
Stores the density level output from the comparator 11.
The comparator 12 compares the input image data with the minimum density level stored in the minimum density memory 14 and outputs the smaller density level. The minimum density memory 14 stores the minimum density level output from the comparator 12.

The maximum density level and the maximum density level output by the maximum density memory 13 and the minimum density memory 14 updated by inputting all pixel image data of the first predetermined number of lines of the original to the comparators 11 and 12, respectively. Minimum density level is D range (dynamic range) detection circuit 1
5 is supplied. The D range detection circuit 15 detects the D range from the maximum density level and the minimum density level, and supplies the D range to the density correction operation circuit 16. The density correction operation circuit 16 outputs image data from the maximum density level and the minimum density level output from the maximum density memory 13 and the minimum density memory 14, respectively, and the D range output from the detection circuit 15.
A density correction coefficient multiplied by the data is calculated, a corrected pixel density corresponding to the input pixel density is obtained based on the obtained density correction coefficient, and this is output to the density correction table 17.

The density correction table 17 stores the input pixel density and the corresponding corrected pixel density as a pair. The switch 18 receives a high-level signal output when the density correction table 17 has completed storing the corrected pixel density, and
It becomes N. The density correction circuit 19 uses the density correction table 17 to perform density correction on the pixels on the lines following the first predetermined number of lines of the document input via the switch 18. The image data subjected to the density correction processing by the circuit 19 is output to the area identification circuit 20. Area identification circuit 20
Extracts a feature amount from the input image data after the density correction, identifies whether the image to be processed belongs to a binary image area or a halftone area, and outputs an identification result.

In the above embodiment, the characteristic amount is detected from the first predetermined number of lines of the document, but the characteristic amount may be detected for at least one line.

Although it is not always necessary to detect the feature amount from the first line, the density of the background can be easily detected since characters are not included in the first line. Further, the area of the line following the first line can be easily identified thereafter.

Further, in the above embodiment, the binarization processing is not performed on the pixels of the first predetermined number of lines used for detecting the characteristic amount. This is because the first predetermined number of lines generally do not include characters and line drawings, so that there is no need to perform binarization processing. However, when a character or a line drawing is included, it is better to perform the binarization processing.

[0038]

[0039]

[0040]

Effects of the Invention] According to the image region identification method of the image region identification apparatus and claim 2 of claim 1, document, or a character and halftone photograph of a mixed image, after it is determined whether a character and a photograph of the mixed image Since the region identification is performed, the identification can be performed more accurately.

[0041]

[0042]

[0043]

[0044]

[0045]

[0046]

[0047]

[0048]

[Brief description of the drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of an image area identification apparatus according to the present invention.

FIG. 2 is a graph showing an example of a histogram detected by the embodiment of FIG.

FIG. 3 is a block diagram showing the configuration of another embodiment of the image area identification device of the present invention.

[Explanation of symbols]

 3 Background density detection circuit 4 Threshold selector 7 Original mode determination circuit 11, 12 Comparator 13 Maximum density memory 14 Minimum density memory 15 D range detection circuit 17 Density correction circuit C1, C2 ... Cn counter

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Toru Fujii 10th Hanazono Todocho, Ukyo-ku, Kyoto-shi, Kyoto Prefecture (72) Inventor Hirohide Ushida 10th Hanazono-Todocho, Ukyo-ku, Kyoto-shi, Kyoto OMRON Corporation (72) Inventor Hajime Okumura 10 Okado Dodocho, Ukyo-ku, Kyoto City, Kyoto Prefecture (56) References JP-A-1-141469 (JP, A) JP-A-63-40461 (JP, A) JP-A-1-41571 (JP, A) JP-A-3-54680 (JP, A) JP-A-3-153167 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) H04N 1/40-1/409 H04N 1/46 H04N 1/60

Claims (2)

(57) [Claims] 1. An image area identifying apparatus for identifying an area of a processing target image area to which of a plurality of different image areas belongs , wherein a density level of pixels of at least one line of a document including the processing target image is included. detecting a feature amount from the detecting means and said detected the document from the feature value letters and Amiten'utsushi
Determine if it is a true mixed image or a mixed image of text and photos
Document mode - a de determination means, the region identified using the output of the original mode judging means
An image region identification apparatus characterized by performing the following . 2. An image area identifying method for identifying an area of a processing target image area to which of a plurality of different image areas belongs , wherein a density level of pixels of at least one line of a document including the processing target image is included. Detection step for detecting feature values from
If the document from the detected feature quantity character and Amiten'utsushi
Determine if it is a true mixed image or a mixed image of text and photos
Document mode - and a de determination step, said area by using an output in said document mode determination step
An image region identification method characterized by performing identification.