JP2011112810A - Image processing method and apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing method and apparatus, capable of determining the degree of focusing in consideration of varying the degree of focusing with the brightness of a lighting device or the like. <P>SOLUTION: The method includes: a second derivative image forming step in which a second derivative image is formed by differentiating twice the respective pixels in the region of objects of decision including a display of object of reading and its environmental background; an absolute mean value calculation process in which an absolute mean value is derived by arithmetically averaging the absolute values of the second derivative value of the respective pixels in the second derivative image: a density histogram forming process in which a density histogram of the region of objects of decision is formed; a density distribution width calculation process in which a threshold for classifying the density histogram into a high luminance side and a low luminance side is derived, and the density distribution width is calculated by the difference of the respective mean densities of the two classes classified by the threshold; a parameter calculation process in which the focusing degree is calculated by dividing the absolute mean value by density distribution width; and a parameter decision process in which the derived focusing degree and a predetermined threshold are compared. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an image processing method and an image processing apparatus, and more particularly to an image processing method and an image processing apparatus for determining the degree of focus of an optical system that reads a reading target display or the sharpness of the reading target display.

  In recent years, with the widespread use of FA (Factory Automation), in a factory production line, an industrial product conveyed by a belt conveyor or the like is imaged using an imaging device such as a camera, and the captured image is processed. Thus, a code symbol such as a bar code symbol and a two-dimensional code symbol, and a reading target display such as a character are automatically read from the industrial product. As described above, when the reading target display attached to the moving subject is read based on the image picked up by the camera fixed at a predetermined position, since there is only one photo opportunity, each subject to be picked up The image is not captured by the autofocus method for obtaining the optimum focal length for the lens, but the focus (focal point) and aperture of the lens provided in the imaging device, the shutter speed of the imaging device, and the object to be imaged are irradiated. Imaging is performed by adjusting in advance imaging conditions such as the amount of irradiation light of the illumination device (see, for example, Patent Document 1).

  When reading a target display by image processing, the contrast between the display area corresponding to the target display and the background area around the display area, and the degree of out-of-focus, greatly affect the reading accuracy. Is generally known. Therefore, when adjusting the imaging conditions in advance as described above, it is determined whether the contrast and focus are appropriate by operating the production line on a trial basis and imaging a test object. Has been done. Conventionally, as a determination item for determining contrast and focus, evaluation using a degree of focus representing the degree of focus has been performed.

  For example, in Non-Patent Document 1, eight types of methods for determining focus (specifically, “method using Fourier transform”, “method using gradient”, “method based on high frequency amount”, “high luminance”) "Method based on quantity", "Method using entropy", "Method using frequency slope", "Method using dispersion" and "Method using SMD (Sum-Modulus-Difference)") Each method is evaluated using (SEM: Scanning Electron Microscope) images. According to the evaluation in Non-Patent Document 1, among the eight types of methods, the “method based on the amount of high frequency” and the “method using dispersion” are excellent (the detection rate of the focused image is 100%). It is shown. Patent Document 2 discloses an image processing method for extracting a rectangular region including a degree of focus in an image.

Patent No. 4080897 JP 2001-331806 A

ViEW2003 Vision Technology Practical Workshop Lecture Collection pp. 12-17

  The focus and contrast appropriateness (focusing degree) of the reading target display such as code symbols and characters attached to the subject is determined by the adjustment process of the optical system such as the imaging device and the illumination device and the reading target display attached to the subject. Used in the inspection process. As described in Non-Patent Document 1, in the method of secondarily differentiating each pixel of a captured image and calculating the degree of focus based only on the second derivative value, brightness variation such as illumination is affected. Along with this, the degree of focus changes. Therefore, the condition for determining the degree of focus is not satisfactory.

  An object of the present invention is to provide an image processing method and an image processing apparatus capable of performing a focus degree determination in consideration of a change in a focus degree depending on brightness such as illumination.

The present invention is an image processing method for determining the degree of focus of an optical system that reads a reading target display or the sharpness of the reading target display by imaging an imaging target region including a predetermined reading target display.
A secondary differential image generating step of generating a secondary differential image by performing secondary differentiation on each pixel in a determination target region including a reading target display and a background around the reading target display;
An absolute average value calculating step of calculating an absolute average value by arithmetically averaging the absolute value of the second derivative value of each pixel in the second derivative image;
A density histogram generating step for generating a density histogram of the determination target region;
A density distribution that calculates a threshold value for dividing the density histogram into a high-luminance class and a low-luminance class, and calculates a density distribution width based on the difference between the average densities of the two classes divided by the threshold value Width calculation step;
A parameter calculating step for calculating a focus degree or a sharpness by dividing the absolute average value by the density distribution width;
An image processing method comprising: a parameter determination step for comparing the calculated degree of focus or definition with a predetermined threshold value.

  The present invention further includes a zero-cross pixel extracting step of extracting only pixels contributing to zero-cross from each pixel of the secondary differential image generated in the second-order differential image generating step.

The present invention is also an image processing apparatus for determining an in-focus level of an optical system that reads a reading target display or a sharpness of the reading target display by imaging an imaging target region including a predetermined reading target display. ,
A secondary differential image generation unit that generates a secondary differential image by performing secondary differentiation on each pixel in a determination target region including a reading target display and a background around the reading target display;
An absolute average value calculating unit that calculates an absolute average value by arithmetically averaging the absolute values of the second derivative values of each pixel in the second derivative image;
A density histogram generator for generating a density histogram of the determination target region;
A density distribution that calculates a threshold value for dividing the density histogram into a high-luminance class and a low-luminance class, and calculates a density distribution width based on the difference between the average densities of the two classes divided by the threshold value A width calculator;
A parameter calculation unit for calculating a focus degree or a sharpness by dividing the absolute average value by the density distribution width;
The image processing apparatus includes a parameter determination unit that compares the calculated degree of focus or definition with a predetermined threshold value.

  According to the present invention, when determining the suitability of the environment setting of the optical system, it is not determined using a value calculated only from the secondary differential image of the determination target region, but the secondary differential image of the determination target region. Is determined by using the degree of focus obtained by dividing the absolute average value by the density distribution width of the determination target region, that is, by using the normalized degree of focus. The degree of focus can be determined in consideration of the change in focus, and the environment of the optical system can be set so that the reading target display can be read with high accuracy. Further, even if the user who sets the environment of the optical system may be different, variations in reading accuracy can be reduced.

  In addition, when inspecting a reading target display such as characters attached to a subject, it is not determined using a value calculated only from the secondary differential image of the determination target candidate area, but the secondary differential image of the determination target area Is determined using the sharpness obtained by dividing the absolute average value by the density distribution width of the determination target area, that is, it is determined using the normalized sharpness. Therefore, it is possible to determine the sharpness in consideration of fluctuations in the image quality, and to accurately detect an abnormality with respect to the reading target display such as printed characters. Moreover, even if the user who sets the environment of the optical system may be different, variation in detection accuracy can be reduced.

1 is a block diagram showing a simplified configuration of an image processing system 1 including an image processing device 20 according to an embodiment of the present invention. 4 is a flowchart of optical system environment setting performed in advance for the imaging device 10 and the illumination device 30. It is a flowchart of the determination process of the adjustment result based on the focus degree in step s3 of FIG. It is a block diagram which shows the functional structure of the image processing apparatus 20 which concerns on this embodiment. It is a figure which shows the Laplacian filter of 3x3 pixel which is a secondary differential filter. It is a figure which shows an example of a density histogram. It is a block diagram which shows the functional structure of the image processing apparatus 20a which concerns on other embodiment of this invention. 6 is a flowchart of an inspection process for inspecting characters attached to a subject 2. It is a flowchart of the process of calculating the density distribution width in step a3 of FIG. It is a flowchart of the calculation process of the absolute average value in step a4 of FIG.

  FIG. 1 is a block diagram showing a simplified configuration of an image processing system 1 including an image processing apparatus 20 according to an embodiment of the present invention. The image processing system 1 captures an imaging target region including code symbols such as barcode symbols and two-dimensional code symbols attached to a subject 2 and characters (including symbols), and outputs the imaging target region as digital image data. 10, an image processing device 20 that performs predetermined image processing on digital image data output from the imaging device 10, and an illumination device 30 that irradiates the subject 2 when the imaging device 10 captures an imaging target region A display device 40 that has a display screen and displays an image of an imaging target region imaged by the imaging device 10, and optical system environment settings (imaging condition setting) performed for the imaging device 10 and the illumination device 30 For example, an input device 50 for inputting a set value or the like is included.

  The image processing system 1 having such a configuration is provided adjacent to a production line of a factory, for example, for automatically reading a code symbol and characters attached to an industrial product that is a subject 2, or an industrial product. In order to check whether the code symbols and characters attached to are properly displayed, they can be suitably used. Hereinafter, a display to be read using the image processing system 1 and a display to be inspected using the image processing system 1 such as code symbols and characters attached to the subject 2 are collectively referred to as a reading target display. It may be noted.

Hereinafter, each device constituting the image processing system 1 will be described in detail.
The imaging device 10 is output from a CCD camera 11 including an optical lens and a charge coupled device (CCD) that outputs an amount of light received by the optical lens as an analog image signal. A / D (Analog / Digital) converter 12 that converts the analog image signal into digital data and outputs it as digital image data, a frame memory 13 that can store digital image data for each frame, and digital image data Each frame is stored in the frame memory 13, and the digital image data is output to a D / A (Digital / Analog) converter 15 for displaying an image on the display device 40, and output from the camera controller 14. The converted digital image data is converted into an analog image signal corresponding to the display device 40. And a D / A converter 15 that outputs to the display device 40.

  Although not shown, the CCD camera 11 has a focus mechanism for adjusting the focus of the optical lens, an aperture mechanism for adjusting the aperture, and an exposure time during which the CCD is exposed to light passing through the optical lens, that is, A shutter mechanism for adjusting the shutter speed is further provided. Each of these mechanisms is configured to be automatically driven by a control signal from the CPU 21, which will be described later, based on a command or set value input from the input device 50, and includes an adjustment knob. It is configured to be driven by being manually displaced.

  The display device 40 is realized by an LCD (Liquid Crystal Display), a CRT (Cathode Ray Tube) display, or the like, and displays an image on a display screen based on an analog image signal output from the D / A converter 15.

  The input device 50 is realized by a pointing device such as a mouse or a keyboard, and is operated by a user to input commands and setting values.

  The image processing apparatus 20 includes a CPU (Central Processing Unit) 21, a RAM (Random Access Memory) 22, a ROM (Read Only Memory) 23, and an I / O (Input / Output) controller 24 as hardware. Consists of. Based on the image processing program stored in the ROM 23, the CPU 21 executes predetermined image processing for performing focus degree determination or sharpness determination described later on the digital image data output from the imaging device 10. At the same time, the overall operation of the image processing system 1 is controlled based on a control program stored in the ROM 23. The image data being processed and the data being calculated are temporarily stored in the RAM 22. The I / O controller 24 is connected to the input device 50 and the power supply 32 of the lighting device 30 and controls these input / output data.

  The illumination device 30 includes a light source 31 for irradiating the subject 2 and a power source 32 that supplies power to the light source 31. The power source 32 is configured to be capable of adjusting the amount of power supplied to the light source 31 and is configured to be automatically adjusted by a control signal from the CPU 21 based on a command or setting value input from the input device 50. In addition, an adjustment knob is provided, and the knob is adjusted by manually displacing the knob. In addition, for example, a light emitting diode (LED) or a tungsten lamp is preferably used as the light source 31, and emits light with an irradiation light amount corresponding to the amount of power supplied from the power supply 32.

  As described above, the image processing system 1 including the image processing apparatus 20 sequentially reads the reading target display attached to each subject (industrial product) 2 being transferred by a conveying unit such as a belt conveyor in a production line of a factory. Or a system for inspecting a reading target display. Therefore, it is not configured to take an image using an autofocus method that automatically sets an optimum focal length for each subject 2, but is configured to set the optical system environment, that is, focus adjustment and aperture adjustment in the CCD camera 11. After the shutter speed adjustment and the irradiation light amount adjustment in the illumination device 30 are performed in advance, an imaging target region including a reading target display is imaged for each subject 2. The image processing system 1 may be used to read or inspect a reading target display attached to an industrial product stopped at a predetermined position.

  Hereinafter, with reference to FIG. 2 to FIG. 4, environment setting of the optical system using the image processing apparatus 20 according to the present embodiment will be described.

  FIG. 2 is a flowchart of the optical system environment setting performed in advance for the imaging device 10 and the illumination device 30. When the preparation for setting the environment of the optical system is completed in step s0, the process proceeds to step s1. In step s1, the shutter speed of the CCD camera 11 is set by the following method, for example. First, the user inputs via the input device 50 whether the subject 2 to be imaged is captured in a stationary state or in a moving state. When the image is captured in a stationary state, the shutter speed is further input to set the shutter speed by driving the shutter mechanism. For example, 1/60 second or 1/100 second is preferably used as the shutter speed when capturing an image of a still object.

In the case of imaging in a moving state, by further inputting the moving speed of the subject 2, the shutter speed is calculated based on the moving speed, and the shutter mechanism is driven based on the calculated value, thereby releasing the shutter. Speed is set. Specifically, the resolution and the allowable movement amount corresponding to the imaging screen area of the imaging device 10 are read from the specifications of the image processing system 1 stored in advance in the RAM 22, and the shutter speed is determined by the following equation (1). Calculated.
Shutter speed = movement speed / (resolution × allowable movement) (1)

When the shutter speed is set, the process proceeds to step s2.
In step s2, an imaging target area including a processing result when image processing is performed on digital image data obtained by imaging the test subject 2 with the reading target display, and a reading target display displayed on the display screen. On the basis of visual observation of the image, the focus mechanism and the diaphragm mechanism are driven by driving the focus mechanism and the diaphragm mechanism in the CCD camera 11, and the amount of power supplied to the light source 31 is adjusted, and the amount of irradiation light is adjusted. . The focus, aperture, and irradiation light quantity are set to desired setting values by adjustment, and when these settings are completed, the process proceeds to step s3.

  In step s3, the suitability of the adjustment result of the focus, stop, and irradiation light amount adjusted in step s2 is determined based on the degree of focus that is a parameter indicating the degree of focus. Details of the focus degree determination will be described later. If it is determined that the adjustment result is appropriate, the process proceeds to step s4. If it is determined that the adjustment result is inappropriate, it is determined that the adjustment result is appropriate. The focus, aperture, and irradiation light amount are adjusted repeatedly until In step s4, the environment setting of the optical system is completed.

  FIG. 3 is a flowchart of the adjustment result determination process based on the degree of focus in step s3 of FIG. FIG. 4 is a block diagram showing a functional configuration of the image processing apparatus 20 according to the present embodiment.

  The image processing apparatus 20 includes an approximate position detection unit 61, a secondary differential image generation unit 62, an absolute average value calculation unit 63, a density histogram generation unit 64, a density distribution width calculation unit 65, and a focus degree calculation unit. 66 and an in-focus determination unit 67. Each of these units 61 to 67 is realized by the CPU 21 executing an image processing program stored in the ROM 23. The functions of the units 61 to 67 will be described together with the description of the adjustment result determination process based on the degree of focus.

  In step s30, the preparation for imaging the test subject 2 is completed using the CCD camera 11 and the illumination device 30 in which the focus, aperture, and amount of irradiation light are set to desired setting values in step s2 of FIG. Then, the process proceeds to step s31.

  In step s <b> 31, the test subject 2 is imaged by the imaging device 10. At this time, when the subject 2 is imaged in a moving state on the actual production line, the test subject 2 is also imaged while moving at the same moving speed, and the subject 2 is placed at a predetermined position on the actual production line. When imaging is performed in a stopped state, the test subject 2 is also imaged in a stopped state at the same position. When the imaging target area including the reading target display in the test subject 2 is captured by the imaging apparatus 10, and the digital image data corresponding to the imaging target area is output from the imaging apparatus 10, the process proceeds to step s32.

  In step s32, the approximate position detection unit 61 of the image processing device 20 detects the approximate position of the reading target display attached to the subject 2 based on the digital image data generated by the imaging device 10. The approximate position detection method can be performed by a well-known method, and can be realized by, for example, OCR (Optical Character Recognition) when the reading target display is a character. When the reading target display is a code symbol, it can be realized by a labeling method, an edge direction histogram method, or the like. Further, when a plurality of characters are included in the imaging target area, an approximate position is detected for each character, and when a code symbol and a character are included in the imaging target area, The approximate position is detected. When the approximate position is detected, the process proceeds to step s33.

  In step s33, one code symbol or character whose approximate position is detected by the approximate position detection unit 61 from the digital image data corresponding to the imaging target area by the secondary differential image generation unit 62 of the image processing device 20, and Image data of a determination target area including a code symbol or a background around a character is extracted. In the following description, it is assumed that the image data of the determination target area is image data of a rectangular area having a size of m × n (where m and n are positive integers) pixels.

  When the image data of the determination target area is extracted, each pixel in the determination target area is second-order differentiated using a second-order differential filter stored in advance in the RAM 22. FIG. 5 shows a 3 × 3 pixel Laplacian filter which is a second order differential filter. FIG. 5A shows a 4-neighbor Laplacian filter LF1, and FIGS. 5B and 5C show 8-neighbor Laplacian filters LF2 and LF3, respectively. By performing the second order differentiation using a Laplacian filter as shown in FIG. 5, a second order differential image composed of each second-order differentiated pixel is generated. When the secondary differential image is generated, the process proceeds to step s34.

In step s34, the absolute value of the secondary differential value is calculated for each pixel of the m × n pixel secondary differential image generated by the secondary differential image generation unit 62 by the absolute average value calculation unit 63 of the image processing device 20. Further, an absolute average value is calculated by arithmetically averaging the absolute values of the secondary differential values of each pixel. Specifically, the secondary differential value of each pixel is U _ij (where i = 1, 2,..., M−1, m and j = 1, 2,..., N−1, n), When the absolute average value is M, the absolute average value M is calculated by the following equation (2).

When the absolute average value M is calculated, the process proceeds to step s35.
In step s35, the density histogram generating unit 64 of the image processing apparatus 20 generates a density histogram representing the density frequency as a histogram for the determination target region of m × n pixels. FIG. 6 is a diagram illustrating an example of a density histogram. In the present embodiment, as described above, the determination target area includes a display area corresponding to the reading target display and a background area corresponding to the background. Therefore, a bimodal density histogram as shown in FIG. Generated. When the density histogram is generated, the process proceeds to step s36.

  In step s36, the density distribution width for the determination target region of m × n pixels is calculated by the density distribution width calculation unit 65 of the image processing apparatus 20 based on the density histogram generated by the density histogram generation unit 64.

  Specifically, first, by using a discriminant analysis method, the density value kmax is determined such that the variance between the two classes is maximized when the density histogram is divided into two classes using a certain density value k as a threshold value. To do. Hereinafter, this density value kmax will be referred to as a binarization threshold value.

For example, when the density range is 0 to D (for example, D = 255), the class having the density range of 0 to k is C0, and the class having the density range k + 1 to D is C1. Further, it is assumed that the frequency for each density value X in the density range, that is, the number of pixels is H (X). At this time, power averaged u0 class C0, w0 the probability of class C0, power averaged u1 class C1, the probability of class C1 w1, the variance between the classes C0 and class C1 When ^{S 2} The dispersion S ² is obtained as a function of the density value k by the following equation (3).
S ² = w0 × w1 × (u1-u0) ² (3)

However,

This dispersion S ^2, by changing the density values k, namely by calculating for each density value k to 0 to D, the value of the variance S ² for each concentration value k is calculated. Thus among the calculated variance S ² and the concentration value k when the variance S ² is maximized, it is determined as binarization threshold kmax.

When the binarization threshold value kmax is determined in this manner, the binarization threshold value kmax causes the density histogram to be classified into the class C0 ′ on the low luminance side (density range is 0 to kmax) and the high luminance side (density range is kmax + 1). ˜D) class C1 ′. At this time, when the average density of class C0 ′ is f0, the average density of class C1 ′ is f1, and the density distribution width is B, the density distribution width B is calculated by the following equation (4).
B = f1-f0 (4)

However,

When the density distribution width B is calculated, the process proceeds to step s37.
In step s 37, the absolute average value M calculated by the absolute average value calculation unit 63 is divided by the density distribution width B calculated by the density distribution width calculation unit 65 by the focus degree calculation unit 66 of the image processing apparatus 20. Thus, the degree of focus Q representing the degree of focus is calculated. Specifically, the degree of focus Q is calculated by the following equation (5).
Q = M ÷ B × 10000 (5)

When the focus level Q is calculated, the process proceeds to step s38.
In step s38, the focus determination unit 67 of the image processing apparatus 20 performs the focus degree determination based on the focus degree Q calculated by the focus degree calculation unit 66. Specifically, the calculated focus degree Q is compared with a focus degree determination threshold value Q0 that is set in advance and stored in the RAM 22, and when it is determined that Q ≧ Q0, the focus is obtained. If it is determined that the degree Q is appropriate, the process proceeds to step s40. If it is determined that Q <Q0, the degree of focus Q is determined to be inappropriate and the process proceeds to step s39. As the threshold value Q0 for determining the degree of focus, for example, a value obtained by an empirical rule is set in advance, and the value is stored in the RAM 22 when the user inputs the value via the input device 50.

  In step s39, as in step s2 in FIG. 2, for example, the processing result when the digital image data obtained by imaging the test subject 2 is image-processed, and the reading target display displayed on the display screen are included. Based on visual observation of the image of the imaging target region, the focus mechanism and the diaphragm mechanism in the CCD camera 11 are driven again to perform focus adjustment and diaphragm adjustment, and irradiation is performed by adjusting the amount of power supplied to the light source 31. Adjust the light intensity. When the desired setting values of focus, aperture, and irradiation light quantity are set by adjustment, the process proceeds to step s31.

  As described above, the focus, the aperture, and the amount of irradiation light are adjusted until it is determined that the focus degree Q is appropriate in the focus degree determination, and when it is determined that the focus degree Q is appropriate, step s40. Then, the adjustment result determination step based on the degree of focus is completed.

  As described above, according to the present embodiment, when determining the suitability of the environment setting of the optical system, the determination target region is not determined using a value calculated only from the secondary differential image of the determination target region. Is determined using the focus degree Q obtained by dividing the absolute average value M of the secondary differential image by the density distribution width B of the determination target region, that is, the determination is performed using the normalized focus degree Q. Therefore, it is possible to determine the degree of focus considering that the degree of focus varies depending on the brightness of illumination, etc., and to set the environment of the optical system so that the reading target display can be read accurately. Can do. Further, even if the user who sets the environment of the optical system may be different, variations in reading accuracy can be reduced.

  Next, an image processing system 1 including an image processing device 20a according to another embodiment will be described. Since the configuration of the image processing system 1 is configured by the same device configuration as that of the above-described image processing system 1 shown in FIG. 1, overlapping description of each device is omitted. The image processing system 1 including the image processing apparatus 20a according to the present embodiment is different from the above-described image processing system 1 in the functional configuration of the image processing apparatus realized by executing the image processing program. In the following, the same components as those of the image processing system 1 described above will be described with the same reference numerals.

  FIG. 7 is a block diagram showing a functional configuration of the image processing apparatus 20a according to the present embodiment. The image processing apparatus 20a includes an approximate position detection unit 71, an inspection target mask image generation unit 72, a density distribution width detection target image generation unit 73, a density histogram generation unit 74, a density distribution width calculation unit 75, a secondary A differential image generation unit 76, a sharpness detection candidate image generation unit 77, a sharpness detection image generation unit 78, an absolute average value calculation unit 79, a sharpness calculation unit 80, and a sharpness determination unit 81 are included. Composed. Each of these units 71 to 81 is realized by the CPU 21 executing an image processing program stored in the ROM 23. The functions of the units 71 to 81 will be described together with the description of the inspection process of the reading target display described later.

  The image processing system 1 including the image processing apparatus 20a according to the present embodiment inspects whether or not the reading target display attached to the subject (industrial product) 2 transferred by a conveying means such as a belt conveyor is properly displayed. In particular, the system is suitable for inspecting whether the display is properly performed.

  Hereinafter, the inspection process of the reading target display using the image processing apparatus 20a according to the present embodiment will be described with reference to FIGS. 7 to 10 on the assumption that the reading target display is a character.

  FIG. 8 is a flowchart of an inspection process for inspecting characters attached to the subject 2. In step a0, when the CCD camera 11 and the illumination device 30 in which the focus, the aperture, and the amount of irradiation light are set to predetermined set values are used, the preparation for imaging the subject 2 with the character to be inspected is completed. The process proceeds to step a1.

  In step a1, the subject 2 to be inspected is imaged by the imaging device 10. When an imaging target area including characters in the subject 2 to be inspected is captured by the imaging apparatus 10 and digital image data corresponding to the imaging target area is output from the imaging apparatus 10, the process proceeds to step a2.

  In step a2, the approximate position of the character attached to the subject 2 is detected by the approximate position detector 71 of the image processing device 20a based on the digital image data generated by the imaging device 10. The approximate position detection method can be performed by a well-known method as described above, and can be realized by, for example, the OCR method. When a plurality of characters are included in the imaging target area, the approximate position is detected for each character. When the approximate position is detected, the process proceeds to step a3.

  In step a3, the density distribution width is calculated. FIG. 9 is a flowchart of the process of calculating the density distribution width in step a3 of FIG. In step a30, processing for calculating the density distribution width is started, and the process proceeds to step a31.

  In step a31, one character whose approximate position is detected by the approximate position detection unit 71 from the digital image data corresponding to the imaging target region by the inspection target mask image generation unit 72 of the image processing apparatus 20a, and the surroundings of the character The image data of the determination target candidate area including the background of the image is extracted. In the following description, it is assumed that the image data of the determination target candidate area is image data of a rectangular area having a size of m × n (where m and n are positive integers) pixels.

  When the image data of the determination target candidate area is extracted, binarization processing is further performed for each pixel in the determination target candidate area. At this time, the binarization process is performed by setting the character to white and the background to black.

  The threshold value for the binarization process may be given by a fixed numerical value stored in advance in the RAM 22 when the user inputs it via the input device 50, and is calculated by using the above-described discriminant analysis method. May be given by the binarization threshold kmax. By performing the binarization process, a binarized image composed of binarized pixels is generated. When the binarized image is generated, the process proceeds to step a32.

  In step a32, the inspection target mask image generation unit 72 performs an expansion process on the binarized image of the determination target candidate area generated in step a31. At this time, the number of times the expansion process is performed (the number of expansions) N1 is stored in advance in the RAM 22 when the user inputs it via the input device 50. The number of expansions N1 is, for example, 2 times. When an expansion image is generated by performing the expansion process, the process proceeds to step a33.

  In step a33, the inspection target mask image generation unit 72 performs a contraction process on the binarized image of the determination target candidate area generated in step a31. At this time, the number of times the contraction process is performed (the number of contractions) N2 is stored in advance in the RAM 22 when the user inputs it via the input device 50. The number of contractions N2 is, for example, 2 times. When a contraction image is generated by performing the contraction process, the process proceeds to step a34.

In step a34, the inspection object mask image generation unit 72 calculates the difference between the expanded image generated in step a32 and the contracted image generated in step a33, thereby corresponding to the contour region of the character to be inspected. A contour image is generated. Hereinafter, this contour image may be referred to as a mask image. Specifically, a mask image is generated by the following equation (6).
Mask image = expanded image-contracted image (6)

When the mask image is generated, the process proceeds to step a35.
In step a35, the density distribution width detection target image generation unit 73 of the image processing apparatus 20a performs a logical product operation on the determination target candidate region image extracted in step a31 and the mask image generated in step a34. Thus, a density distribution width detection target image is generated. This density distribution width detection target image corresponds to an image obtained by extracting pixels in the contour region of the inspection target character, which is the determination target region, from the determination target candidate region image. Specifically, the density distribution width detection target image is generated by the following equation (7).
Density distribution width detection target image = candidate area image & mask image (7)

When the density distribution width detection target image is generated, the process proceeds to step a36.
In step a36, the density histogram generation unit 74 of the image processing apparatus 20a generates a density histogram for the density distribution width detection target image, that is, a determination target region. When the density histogram is generated, the process proceeds to step a37.

  In step a37, the density distribution width for the determination target region is calculated by the density distribution width calculation unit 75 of the image processing apparatus 20a based on the density histogram generated by the density histogram generation unit 74. Specifically, first, as described above, the binarization threshold value kmax is determined by using the discriminant analysis method. When the binarization threshold kmax is determined, and the density histogram is divided into the low luminance side class C0 ′ and the high luminance side class C1 ′ by the binarization threshold kmmax, the average of the class C0 ′ is determined. The density distribution width B is calculated by the above equation (4), where the density is f0, the average density of class C1 ′ is f1, and the density distribution width is B. When the density distribution width B is calculated, the process proceeds to step a38. In step a38, the process for calculating the density distribution width B ends. Returning to FIG. 8 again, when the density distribution width B is calculated, the process proceeds to step a4.

  In step a4, a secondary differential image of the determination target region (character contour region) in the determination target candidate region of m × n pixels is generated, and an absolute average value is calculated for the secondary differential image. FIG. 10 is a flowchart of the absolute average value calculation process in step a4 of FIG. In step a40, processing for calculating an absolute average value is started, and the process proceeds to step a41.

  In step a41, a smoothing filter, that is, an average value filter or a Gaussian filter is applied to the image data of the determination target candidate region extracted by the inspection target mask image generation unit 72 by the secondary differential image generation unit 76 of the image processing apparatus 20a. By using (weighted average filter), smoothing processing is performed and noise is removed. When the smoothing process is performed, the process proceeds to step a42.

  In step a42, each pixel of the image data of the determination target candidate area that has been subjected to the smoothing process by the second-order differential image generation unit 76 is converted into a pixel value using a Laplacian filter stored in advance in the RAM 22, as described above. Second derivative. When a secondary differential image composed of each pixel subjected to the secondary differentiation is generated, the process proceeds to step a43.

In step a43, the second differential image of the determination target candidate region generated by the secondary differential image generation unit 76 and the inspection target mask image generation unit 72 are generated by the definition detection candidate image generation unit 77 of the image processing apparatus 20a. A sharpness detection candidate image in which each pixel in the determination target region is second-order differentiated is generated by performing a logical product operation with the mask image corresponding to the determined determination target region (character outline region). Specifically, a sharpness detection candidate image is generated by the following equation (8).
Sharpness detection candidate image = secondary differential image & mask image of determination target candidate region (8)

When the sharpness detection candidate image is generated, the process proceeds to step a44.
In step a44, the sharpness detection image generation unit 78 of the image processing device 20a extracts pixels contributing to zero crossing from among the pixels of the sharpness detection candidate image generated by the sharpness detection candidate image generation unit 77. As a result, a sharpness detection image is generated. That is, in this embodiment, unnecessary pixels are further removed from each pixel of the secondary differential image (definition detection candidate image) in the determination target region, thereby obtaining a secondary differential image (clear) corresponding to the new determination target region. Degree detection image) is generated.

  Specifically, for each pixel of the sharpness detection candidate image, the product of the secondary differential value of the pixel of interest and the secondary differential value of each pixel adjacent to the adjacent 8 adjacent to the pixel of interest is calculated. If at least one of the values (eight integrated values when there are eight adjacent pixels) has a negative integrated value, that pixel of interest becomes the pixel constituting the sharpness detection image, that is, zero cross. On the contrary, if there is no negative integrated value, it is determined that the target pixel is not a pixel constituting the sharpness detection image. In this way, a sharpness detection image is generated by extracting only the pixels that contribute to zero crossing from the pixels of the sharpness detection candidate image. When the sharpness detection image is generated, the process proceeds to step a45.

  In step a45, the absolute value of the secondary differential value is determined for each pixel of the sharpness detection image generated by the sharpness detection image generation unit 78 by the absolute average value calculation unit 79 of the image processing device 20a, as described above. Further, an absolute average value is calculated by arithmetically averaging the absolute values of the secondary differential values of the respective pixels. When the absolute average value M is calculated, the process proceeds to step a46. In step a46, the process for calculating the absolute average value M ends. Returning to FIG. 8 again, when the absolute average value M is calculated, the process proceeds to step a5.

In step a5, the absolute value M calculated by the absolute average value calculator 79 is divided by the density distribution width B calculated by the density distribution width calculator 75 by the definition calculator 80 of the image processing apparatus 20a. Thus, the sharpness R representing the degree of sharpness of the character to be inspected is calculated. Specifically, the definition R is calculated by the following equation (9).
R = M ÷ B × 10000 (9)

When the sharpness R is calculated, the process proceeds to step a6.
In step a6, the sharpness determination unit 81 of the image processing apparatus 20a performs the sharpness determination based on the sharpness R calculated by the sharpness calculation unit 80. Specifically, the calculated sharpness R is compared with a sharpness determination threshold value R0 that is set in advance and stored in the RAM 22, and when it is determined that R ≧ R0, the sharpness R is If it is determined that R <R0, it is determined that the sharpness R is inappropriate, and the process proceeds to step a38. For example, a value obtained by an empirical rule is determined in advance as the threshold value R0 for determining the sharpness, and is stored in the RAM 22 when the user inputs the value via the input device 50.

  In step a37, since the sharpness R is appropriate, that is, it is determined that there is no abnormality in the printed characters, the inspection process is terminated. Further, in step a38, it is determined that the sharpness R is inappropriate, that is, the printed character is abnormal. Therefore, for example, the user is notified of the abnormality via the display screen, and step a39. Proceed to In step a39, the inspection process is terminated.

  As described above, according to the present embodiment, when the reading target display such as characters attached to the subject 2 is inspected, the determination is made using the value calculated only from the secondary differential image of the determination target candidate region. Rather, since the determination is made using the sharpness R obtained by dividing the absolute average value M of the secondary differential image of the determination target region by the density distribution width B of the determination target region, that is, the normalized definition R Therefore, the sharpness determination can be performed in consideration of the fluctuation of the sharpness depending on the brightness of illumination or the like, and an abnormality with respect to the reading target display such as a printed character can be accurately detected. it can. Moreover, even if the user who sets the environment of the optical system may be different, variation in detection accuracy can be reduced.

  Further, in the present embodiment, pixels that contribute to zero crossing are avoided from the sharpness detection candidate image extracted by the mask image corresponding to the contour region of the reading target display, in order to avoid the influence of a cumulative error of a small second derivative value. Since the sharpness detection image is generated by extracting only and the absolute average value is calculated from this sharpness detection image, the accuracy of the sharpness determination can be improved.

DESCRIPTION OF SYMBOLS 1 Image processing system 10 Imaging device 11 CCD camera 20 Image processing device 21 CPU
22 RAM
23 ROM
DESCRIPTION OF SYMBOLS 30 Illuminating device 31 Light source 32 Power supply 40 Display apparatus 50 Input apparatus 61 Approximate position detection part 62 Secondary differential image generation part 63 Absolute average value calculation part 64 Density histogram generation part 65 Density distribution width calculation part 66 Focus degree calculation part 67 Matching Focus determination unit

Claims (3)

An image processing method for determining an in-focus level of an optical system for reading a reading target display or a sharpness of a reading target display by imaging an imaging target region including a predetermined reading target display,
A secondary differential image generating step of generating a secondary differential image by performing secondary differentiation on each pixel in a determination target region including a reading target display and a background around the reading target display;
An absolute average value calculating step of calculating an absolute average value by arithmetically averaging the absolute value of the second derivative value of each pixel in the second derivative image;
A density histogram generating step for generating a density histogram of the determination target region;
A density distribution that calculates a threshold value for dividing the density histogram into a high-luminance class and a low-luminance class, and calculates a density distribution width based on the difference between the average densities of the two classes divided by the threshold value Width calculation step;
A parameter calculating step for calculating a focus degree or a sharpness by dividing the absolute average value by the density distribution width;
An image processing method, comprising: a parameter determination step of comparing the calculated degree of focus or definition with a predetermined threshold value.   2. The image processing method according to claim 1, further comprising a zero-cross pixel extracting step of extracting only pixels contributing to zero-cross from each pixel of the second-order differential image generated in the second-order differential image generating step. . An image processing apparatus for determining an in-focus level of an optical system for reading a reading target display or a sharpness of a reading target display by imaging an imaging target region including a predetermined reading target display,
A secondary differential image generation unit that generates a secondary differential image by performing secondary differentiation on each pixel in a determination target region including a reading target display and a background around the reading target display;
An absolute average value calculating unit that calculates an absolute average value by arithmetically averaging the absolute values of the second derivative values of each pixel in the second derivative image;
A density histogram generator for generating a density histogram of the determination target region;
A density distribution that calculates a threshold value for dividing the density histogram into a high-luminance class and a low-luminance class, and calculates a density distribution width based on the difference between the average densities of the two classes divided by the threshold value A width calculator;
A parameter calculation unit for calculating a focus degree or a sharpness by dividing the absolute average value by the density distribution width;
An image processing apparatus comprising: a parameter determination unit that compares the calculated degree of focus or definition with a predetermined threshold value.

Images