JPH06103399A - Image processor - Google Patents

Abstract

PURPOSE:To provide the means for discriminating an area of each separate image characteristic at a high speed and with high accuracy by a simpler mechanism than a conventional one, with regard to the image processor which discriminates and separates information areas having different image characteristics such as bar-code data, character data, a photograph, etc., recorded in a document (input business form), and can executes an image processing corresponding to the image characteristic of each area. CONSTITUTION:In the image processor for allowing a document to be subjected to line scan and reading it in a picture element unit, and executing a determined processing with regard to an area having a specific image characteristic in obtained image information, this image processor is constituted so that a neuro- mechanism 12 for inputting a value of a picture element of an arbitrary area is provided, and an area having a specific image characteristic is discriminated and segmented.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention identifies and separates information areas having different image characteristics such as bar code data, character data and photographs recorded on an original (input form), and determines the image characteristics of each area. The present invention relates to an image processing device that enables appropriate image processing. Further, in particular, a means for highly accurately binary-coding the separated barcode data is also provided.

[0002]

2. Description of the Related Art A document on which a plurality of types of data having different image characteristics such as bar codes, characters, dots and photographs are recorded is uniformly read by a CCD or the like, converted into a multivalued video signal, and then each data is converted. When performing image processing according to the above, it is necessary to divide the area of the image of the read result according to the image characteristics of the data.

For example, in the bar code area, the bar code is scanned to detect the width and interval of each bar code, and 2
The process of converting into a value code is performed. In the character and dot data areas, simple binarization processing is performed in pixel units. Then, in the photographic region, the pseudo halftone conversion process is performed by performing binarization by the dither method or the like.

FIG. 10 shows the configuration of a conventional image processing apparatus having an image characteristic-based area identification function. In the figure, 1
Is an original in which plural kinds of data having different image characteristics are recorded. Reference numeral 2 is a CCD for raster-scanning the original 1 and converting it into a video signal. An A / D converter 3 converts a video signal into multi-valued image data in pixel units of each line. An input buffer 4 temporarily holds image data of one line or a plurality of lines. 5 is
This is an area identification unit that examines the image data of the sequential lines input to the input buffer 4 and cuts out areas according to image characteristics. Reference numeral 6 denotes coordinates and image characteristic types (bar code, characters, dot painting,
It is an area management table for registering photographs and the like). 7 is
This is an image file that sequentially stores the image data for each line held in the input buffer 4. Reference numerals 8, 9, and 10 denote image processing units that perform processing according to image characteristics.

Next, the function of the area identifying section 5 will be described. FIG. 11 shows the video signal waveform of the CCD output,
(A) shows the waveform of the barcode area, (b) shows the waveform of the character area, (c) shows the waveform of the dot image, and (d) shows the waveform of the photographic area. As shown in (a), (b), and (c) in the figure, the image of the area of the bar code, the character, and the dot image represented only by the presence / absence of black ink has a wide range of the maximum and minimum densities. The slope (edge of the waveform) is steep, and each has its own periodicity of length. On the other hand, the image in the photographic area has an image characteristic that the difference between the maximum value and the minimum value of the density is relatively small and the density gradient is gentle.

The area discriminating section 5 is a software function for discriminating an image area by utilizing these image characteristics, and detects the difference between the maximum and minimum density values, the density gradient, the periodicity, etc. It consists of several programs that do.
Then, each time image data is written to the input buffer 4, these respective functions are sequentially operated to check the image characteristics,
A region having the same image characteristic is identified, and the coordinate of the region and the image characteristic type are registered in the region management table 6.

When the processing of the area identifying section 5 is completed, the image data of the input buffer 4 is stored in the image data file 7 and the processing of the next line is performed. When all the lines of the original 1 have been processed and the area for each image characteristic is determined, each image processing unit 8, 9, 10 recognizes the area of the image characteristic, which the image processing section 8, 9 and 10 are responsible for, from the area management table 6. Only the area having the corresponding image characteristic is extracted from the image data file 7 and processed. For example, the image processing unit 8 performs the encoding process on the barcode in the barcode region, the image processing unit 9 performs the binarization process on the character region, and the image processing unit 10 performs the binarization process on the photographic region. .

[0008]

In the conventional image characteristic-based area identification function in the image processing apparatus, some programs must be sequentially executed in order to identify the area by detecting the change width and gradient of the density. However, there is a problem that the amount of programs increases and the processing takes time.

An object of the present invention is to provide a means for performing high-speed and high-accuracy region identification for each image characteristic with a mechanism simpler than the conventional one.

[0010]

The present invention is intended to solve the problems by using a neuro mechanism (or a neuro computer) as a means for discriminating regions by image characteristics.

FIG. 1 shows the principle of the present invention. In the figure, 11 is input image data, and the image space corresponding to the document surface is composed of a matrix of pixels having density values.

Reference numeral 12 denotes a neuro mechanism, which sequentially cuts small areas from the input image data 11 to identify image characteristics,
Outputs the image characteristic type. The small area is selected in a size of 8 continuous pixel units on one line, or an 8 × 8 pixel matrix unit on 8 lines. As long as the faithful plot of the CCD output signal waveform is not hindered, the sampled data at an appropriate pixel interval can be input.

Numeral 13 is image characteristic identification data, which is data indicating an image space corresponding to the image space of the input image data 11 divided by image characteristics.

[0014]

In FIG. 1, the input image data 11 is obtained by raster-scanning an original with a CCD or the like and converting the read image signal into a digital signal, and the size of the pixel matrix is determined according to the resolution. . The neuro mechanism 12 is preliminarily learned using sample images for types of image characteristics that require identification. A well-known technique can be used for the structure of the neuro mechanism 12 and the learning method.

The neuro mechanism 12 for discriminating a plurality of types of image characteristics functions as a filter for detecting the learned plurality of types of image characteristics, and is based on pixel values of a small area given by sequentially cutting out from the input image data 11. The image characteristic is determined by outputting the determination result as the image characteristic identification data 13 corresponding to the small area. When the image space of the input image data 11 is completely scanned in a small area and the image characteristic determination is completed, the image characteristic identification data 13 is divided into the same image characteristic area in a map shape as in the illustrated example. Image space can be shown.

The identification process is speeded up by identifying the image characteristics using the neuro mechanism, and the dynamic learning based on the image of the document to be processed is possible. Highly accurate identification performance can be retained.

[0017]

FIG. 2 is a conceptual diagram of an embodiment of the neuromechanism 12 in FIG. 2A shows a neuro structure, which includes an input layer 12a, an intermediate layer 12b, and an output layer 12c. The intermediate layer 15 can be a plurality of layers. The input layer 14 is two-dimensionally arranged so as to have the components of the displacement and the density on the line (main scanning direction) of the input image data, and the information of the density change with respect to the displacement is given.

FIG. 2B shows the output signal waveform of the CCD applied to the input layer 14 of the neuron, and the signals of a plurality of appropriately sampled plot points are applied to the input layer 14 at the same time. In the sampling at this time, the sampling position can be partially changed so that the maximum value and the minimum value of the waveform are always selected.

From the output layer 16, four image characteristic types of bar code, character, dot image and photograph are selectively output as the identified image characteristic. The overall configuration of the embodiment of the present invention is
This can be easily realized by replacing the area identification unit 5 of the conventional device shown in FIG. 10 with the neuro mechanism 12 of FIG. However, a means for sampling the CCD output signal waveform is added after the input buffer 4, and the image characteristic identification data 13 of FIG. 1 is stored in the area management table 6 of FIG.

Next, a bar code reading process will be described as an example of image processing based on the result of the area identification. FIG. 3 is a block diagram of an apparatus according to an embodiment that performs a barcode reading process. In the figure, 1 is a document, 2 is a CCD, 3 is an A / D converter, 6 is an area management table having coordinates of a bar code area, 7 is an image data file, 8 is an image processing unit for reading a bar code, 12 Is a neuro mechanism for area discrimination, 14 is a digital comparator, 15 is a histogram calculation unit, 16
Is a floating slice part, 17 is an average slice part, and 18 is a bar code decoder.

Next, the function of the image processing unit 8 shown in FIG. 3 will be schematically described with reference to FIG. FIG. 4A shows a manuscript 1
It is an example of the bar code recorded in, and shows the state that the width of the bar becomes thick due to the noise of dirt and crushing, or the bars are bridged. FIG. 4B shows the result data due to a reading error when the barcode is scanned and binarized at the reading position shown in FIG. Further, FIG. 4C is correct reading result data when there is no influence of noise as in FIG. 4A.

FIG. 4D shows a vertical histogram created by accumulating the values of pixels in the vertical direction by the histogram calculation unit 15 after the simple binarization process by the digital comparator 14 of FIG. The floating slice executed by the floating slice unit 16 for the histogram is shown. In this case, it can be seen that the reading is possible with the influence of noise considerably reduced.

FIG. 4E shows the histogram calculation unit 15
First, the process of accumulating pixel values in the horizontal direction to create a horizontal histogram and detecting a region affected by noise (shown by hatching) in the average slice unit 17 will be described. FIG. 4F shows a histogram calculation unit 15 which creates vertical histograms by accumulating pixel values in the vertical direction in the remaining area excluding the hatched area shown in FIG. The state where the slicing unit 17 executes the averaging slice is shown. In this case, a more reliable reading result can be obtained.

The bar code decoder 18 selects the slice output of the floating slice unit 16 or the averaging slice unit 17, converts it into a bar code, and outputs it. When the image quality of the original 1 is good, the floating slice section 16 which is fast in processing is used, and when the image quality of the original 1 is bad, the average slice section 17 which is reliable even though the processing time is long is selected.

Details of the floating slice processing will be described with reference to FIGS. 5, 6 and 7. 5 shows a flow of the floating slice processing, FIG. 6 shows an example of a histogram, and FIG. 7 shows a result of the floating slice processing based on the example of the histogram of FIG.

FIG. 5 shows the processing steps of histogram generation in the vertical direction. Here, the X coordinate number is i, the Y coordinate number is j, and v _ij is the pixel value of the coordinate X _i Y _ij (1 /
0) and V _i represent the count value (height of the histogram) of black pixels at the coordinate X _i . First i, the initial value of j is 0, the value of v _ij is added to V _i processing V _i = V
Repeat _i + v _ij until the value of j reaches the maximum value, which is the bar height of the bar code. When j reaches the height of the bar, the value of V _i is stored as a histogram value and i
= I + 1, the X coordinate is advanced by one, the above process is repeated again, and the process proceeds to the floating slice process when i reaches the barcode width.

At the stage of the floating slice processing, the white count and the black count for counting the width of the black and white alternating area of the bar code are used. First, the storage addresses of the white count and the black count are initialized, and the floating slice level A is calculated by the formula shown. Here, n is a work number for determining the slice level calculation target data indicated by 21 in FIG.

If the histogram value V0 of the slice target position when n = 0 is smaller than A, it is determined to be black, and if it is larger, it is determined to be white, and the white count or the black count value is initially 0.
Therefore, in each case, 1 is set, the value of n is incremented by 1 to calculate the slice level A again, V1 is compared with A, and when the result is black, if the white count is not 0, the area from white to black The count value of the immediately preceding white area (white bar width) is stored, and when the comparison result is white, the count value of the immediately preceding black area (black bar width) is similarly stored. Next, add the black or white count corresponding to the current area +
1 and n is updated by +1 to return to the slice level calculation processing, and thereafter, the same processing is repeated until the entire area of the barcode is sliced.

In the example of FIG. 6, 20 histograms are generated from 19 barcode image data, and 21 histograms are generated.
Slice level A is calculated while sequentially selecting slice level calculation target data from _{n−2 to} V _{n + 2} according to n,
White / black determination is performed as in the example shown in FIG. 7, and as a result, the width of the alternating white and black regions is obtained using the count data. Based on this result, barcode decoding processing is performed.

Next, details of the average slice processing will be described with reference to FIGS. 8 and 9. FIG. 8 is a flow of average slice processing, and FIG. 9 shows an example of a histogram. In FIG. 8, 'is a histogram generation processing step in the main scanning direction (horizontal direction),' is a processing step for determining a noise influence area from the histogram generated in ', and'is a vertical histogram in the remaining area excluding the noise influence area. Is a processing step for generating.

The processing of'is similar in operation to the processing of FIG. 5 although there is a horizontal / vertical difference in the generated histogram. In ′, the histogram H _j is calculated, and at the same time, the total SUM of the histograms is calculated, and the histogram average is also prepared. In the process of ', the SUM value is divided by the height of the bar to obtain the histogram average,
When the histogram H _j exceeds this value, the process of setting the flag as being influenced by noise is repeated up to the height of the bar to determine the noise influence region.
In the process of ', the black pixels are counted except for the noise-affected area in which the flag is set, and the histogram in the vertical direction is generated. Except for the points, the processing is the same as that of FIG.

In the example of FIG. 9, scanning direction histograms of 22 are generated for the bar code image data 19 by the processing of ', noise affected areas 23 are detected by the processing of' and vertical processing is performed by the processing of '. A histogram 24 has been generated.

[0033]

As described above, according to the present invention, in the reading of a document in which information of various image characteristics is mixed, it is possible to cut out a region having an arbitrary image characteristic quickly and highly accurately, and to improve the reliability of image processing. Can be increased.

In reading the bar code area,
Even if the image quality is deteriorated by noise, it is possible to read with less error.

[Brief description of drawings]

FIG. 1 is a principle diagram of the present invention.

FIG. 2 is a conceptual diagram of an embodiment of a neuro mechanism.

FIG. 3 is a configuration diagram of an apparatus according to an embodiment that performs a barcode reading process.

FIG. 4 is an explanatory diagram of a bar code reading function of the embodiment apparatus.

FIG. 5 is a flow chart of floating slice processing in the embodiment of the present invention.

FIG. 6 is an explanatory diagram of an example of a histogram of floating slice processing.

FIG. 7 is an explanatory diagram of a floating slice processing result.

FIG. 8 is a flowchart of average slice processing according to the embodiment of the present invention.

FIG. 9 is an explanatory diagram of a histogram example of average slice processing.

FIG. 10 is a block diagram of a conventional image processing apparatus.

FIG. 11 is a video signal waveform diagram of a CCD output.

[Explanation of symbols]

 11 input image data 12 neuro mechanism 13 image characteristic identification data

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

[Procedure amendment]

[Submission date] September 28, 1992

[Procedure Amendment 1]

[Document name to be corrected] Drawing

[Correction target item name] All drawings

[Correction method] Change

[Correction content]

[Figure 1]

[Figure 7]

FIG. 11

[Fig. 2]

[Figure 3]

[Figure 4]

[Figure 5]

[Figure 6]

[Figure 8]

[Figure 9]

[Figure 10]

Claims (3)

[Claims] 1. An image processing apparatus that performs line scanning of an original document to read it in pixel units, and performs a predetermined process on an area having specific image characteristics in the obtained image information. An image processing apparatus, characterized by providing a neuro mechanism as an input to identify and cut out a region having the above-mentioned specific image characteristics. 2. The clipped area according to claim 1, wherein the bar code area is arranged, and the bar code in the bar code area is arranged in parallel with a line, and the pixel value of the bar code area is line by line. To create a horizontal histogram, identify noisy lines from the created histogram values, and for each of the remaining lines excluding the noisy lines, the pixels at the same pixel position on the line. Accumulate values to create a vertical histogram,
An image processing apparatus characterized by binarizing a barcode by slicing the created histogram value with a fixed threshold value. 3. The bar code area according to claim 1, wherein the bar code area is arranged parallel to a line, and each line of the bar code area is on a line. An image processing apparatus characterized in that a histogram in the vertical direction is created by accumulating values of pixels at the same pixel position, and the created histogram is binarized by floating slice.