JPH0636023A - Image signal processor - Google Patents

Abstract

PURPOSE:To obtain the superior image signal processor that removes a noise even in environment of a small quantity of light and is a binarizing device for a digital multivalued image signal which extracts a bar code area from scanned image data like facsimile and decodes it. CONSTITUTION:The image signal processor is provided with a parity noise removing circuit 12 for two-dimensional image data inputted by a raster scan system, a scanning means for a scanning window 13 corresponding to bar code size, a threshold value arithmetic circuit 14, an image emphasizing circuit 15 for a bar code image, and a noise removing circuit 17 which removes noise components other than the bar code from the output of a binarizing and comparing circuit 16; and an image signal of good quality having no noise can be outputted from an image of a subject in the presence of low lighting.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image signal processing device for binarizing a digital multi-valued image signal for extracting and decoding a barcode area from image data scanned like a facsimile.

[0002]

2. Description of the Related Art With the recent development of facsimile machines,
A method of removing noise from image data to obtain high image quality has been proposed.

The conventional image signal processing device will be described below.
A description will be given with reference to the drawings. As shown in FIG. 2, in the conventional image signal processing device, an input image signal (multi-valued) is inputted from an input terminal 1 and binarized by a binarizing circuit 2 and a binarizing circuit which binarizes the image signal at a constant threshold value. A binary data memory 3 for storing the selected signal, a contour detection circuit 4 for outputting the contour data of the binarized signal, and a contour of the bar code by expansion processing for taking the logical sum of the data in the scanning window for scanning the contour data and outputting it. A contour signal selection expansion circuit 5 for filling data, an nxn contraction circuit 6 for outputting a logical product of binary data in an nxn scanning window for scanning selectively expanded data, and a reduction by thinning out expansion data And a contraction / contraction circuit 7 for erasing the background image area while leaving a part of the barcode area by contraction processing for obtaining a logical product of the data in the scanning window for scanning the contracted data and outputting it.
A tilt / center calculation circuit 8 for calculating the tilt and center of the bar code area from the contracted data, an address calculation circuit 9 for reading the binary data memory 3 according to the tilt direction, and a bar for decoding the bar code from the read data of the memory. It is composed of a code decoding circuit 10.

The corner components and their operation will be described below. First, binarization of a multi-valued input image signal is a simple binary method, which is composed of a binarization circuit 2 with a threshold value of a constant level depending on the content of the signal.
No. 84 publication.

[0005]

However, in the above-mentioned conventional configuration, when the subject is out of the depth of focus range or the amount of illumination light to the subject is low, the image signal amplitude including the bar code is small and at the same time there is a lot of noise. A method and apparatus for binarizing only a barcode from a multi-valued image signal have not been published so far.

The present invention solves the above-mentioned conventional problems, and an object of the present invention is to provide an image signal processing device capable of binarizing a bar code having a low light intensity and a high depth of focus.

[0007]

In order to achieve the above object, an image signal processing apparatus of the present invention comprises an even-odd noise removing apparatus for two-dimensional image data input by a raster scan method,
Scanning means of an m × m scanning window according to the barcode size,
threshold calculation means using an m × m scanning window, image enhancement means for a barcode image also using an m × m scanning window, and binarization comparing means binarized by the threshold and the barcode image, And a noise removing means for removing noise components other than the barcode in the binary data.

[0008]

According to the present invention, with the above-mentioned structure, it is possible to binarize a bar code having a low light intensity and a high depth of focus.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. As shown in FIG. 1, in the image signal processing apparatus of the present embodiment, an input image signal (multi-valued) from a CCD linear sensor or the like is input from an input terminal 11, and a linear CC
Even-odd noise removing circuit 12 for removing even-odd noise unique to the D sensor, m × m scanning window 13 according to a barcode size used for processing an input image signal, N points from pixel signal data in scanning window 13 A threshold calculation circuit 14 for obtaining a data average value, and an image enhancement circuit 1 capable of setting a gain for each direction of main scanning and sub-scanning for pixel signal data in the scanning window 13.
5. Binarization comparison circuit 16 that applies a constant level bias to the threshold / image enhancement signal to suppress noise of multilevel signals below the signal and binarizes the binarized signal using the scanning window 13 In addition, it is composed of a noise removing circuit 17 for eliminating isolated noise.

The relationship between the components and the operation thereof will be described below. First, an input image signal (multivalued) from a CCD linear sensor or the like is input from the input terminal 11. A constant level noise N generated every other pixel, which is specific to the CCD sensor, is superimposed on this. Even noise removal circuit 12
Are output pixels Y1 and Y2 from the target pixel X1 including even and odd noise, the target pixel X2 not including even and odd noise, and the preceding and succeeding pixels A and B for the respective target pixels X1 and X2 as follows.

Y1 = {2 × (X1 + N) + A + B} / 4 = (2 × X1 + A + B) / 4 + N / 2 Y2 = {2 × X2 + (A + N) + (B + N)} / 4 = (2 × X2 + A + B) / 4 + N / 2 That is, the noise component N changes to a DC component having an intensity of 1/2 in both the output pixels Y1 and Y2, and the influence on the image signal disappears. The image signal is input to the m × m scanning window 13, and the size of the scanning window 13 is twice the thickness of the thin bar of the barcode used for processing the input image signal. To match. Pixel signals at N points are obtained from the pixel signal data in the scanning window 13 at equal distances in the equidistant direction, and the threshold value calculation circuit 14 calculates the average value of the N point pixels and sets it as the threshold value. The image enhancement circuit 15 can set a variable gain for each direction of main scanning and sub scanning in order to match the difference in MTF with the pixel signal data in the scanning window 13 depending on the difference in scanning direction. The binarization comparison circuit 16 applies a constant level bias to the threshold value, provides a constant difference with respect to the emphasized image signal in the case of a uniform image signal, and obtains stable white / black output during comparison. The noise removing circuit 17 stores the bar code (maintains the linear portion of the bar) and removes the other portion (noise) of the binarized signal using the scanning window 13.

[0012]

As is apparent from the above-described embodiments, according to the present invention, it is possible to binarize a bar code from an image signal attenuated due to a low light amount and a high depth of focus. The signal processing device can be realized.

[Brief description of drawings]

FIG. 1 is a block diagram of an image signal processing apparatus according to an embodiment of the present invention.

FIG. 2 is a block diagram of a conventional image signal processing device.

[Explanation of symbols]

 12 even-odd noise removal circuit 13 scanning window (m × m) 14 threshold value calculation circuit 15 image enhancement circuit 16 binarization comparison circuit 17 noise removal circuit

Claims (1)

[Claims] 1. An even / odd noise removing device for two-dimensional image data input by a raster scan method, a scanning means of an m × m scanning window corresponding to a barcode size, and the m × m scanning window are used. Threshold calculation means, image enhancement means for a barcode image using a scanning window of m × m, binarization comparison means binarized by the threshold value and the barcode image, and a bar in the binary data. An image signal processing apparatus comprising: a noise removing unit that removes noise components other than a code.