JP7080123B2 - Image inspection equipment - Google Patents

Description

The present invention relates to an image inspection device for inspecting the appearance of an object to be inspected.

Conventionally, for example, as disclosed in Patent Document 1, an image inspection device may be used at a site for inspecting the appearance of a work such as various products. The image inspection device performs a search process that detects the position and posture of the work and sets the inspection area, and an inspection process that inspects surface defects such as scratches in the inspection area set based on the result of the search process. Will be. In order to perform the search process, the user sets the pattern area on the set image at the time of setting, and extracts the feature amount of the work in the set pattern area. Then, with respect to the inspection image input during the inspection operation, a region having a high correlation value with the set feature amount is specified, and the position and posture of the work are specified.

Japanese Unexamined Patent Publication No. 2014-55916

By the way, as a method of search processing of an image inspection device, for example, a normalized correlation search is generally often used. The normalized correlation search is a method of searching the pattern region and the image region having the highest normalized correlation value while changing the position and angle of the pattern region. In the case of the normalized correlation search, if there is another region in the inspection image having a shading value similar to the region originally desired to be detected, erroneous detection is likely to occur.

On the other hand, an edge-based search process using an edge in a pattern region as a feature amount is also known. In the edge-based search process, even if a region having a similar shading value exists in the inspection image, if a similar edge does not exist in the inspection image, it does not lead to false detection, so it is normal in this respect. It is advantageous compared to the canonical correlation search. However, in the edge-based search process, if the edge cannot be detected in the inspection image, the search fails.

As described above, the appropriate feature amount used for the search process differs depending on the surface condition and shape of the work. Moreover, even if the search processing is successful with a certain search algorithm, it is difficult for the user to determine how much margin the search processing is successful with and whether there is another more suitable search algorithm. There is a problem.

The present invention has been made in view of this point, and an object of the present invention is to allow a user who does not know the characteristics of a plurality of search algorithms or compatibility with an inspection target in a search process. The purpose is to make it easy to determine the appropriate search algorithm.

In order to achieve the above object, the first invention comprises a registered image storage unit for storing a registered image obtained by imaging the inspection object in an image inspection apparatus for inspecting the appearance of the inspection object, and the above-mentioned. A pattern area setting unit for setting a pattern area used for search processing on the registered image stored in the registered image storage unit, a first feature amount extracted from an image in the pattern area, and the first feature amount. With respect to the feature amount storage unit that stores the second feature amount different from the feature amount and the input image obtained by newly imaging the inspection object, the first feature amount and the predetermined value or more are obtained. A search processing unit that executes a first search process for detecting an image region having a correlation value and a second search process for detecting an image region having a correlation value equal to or higher than a predetermined value with the second feature amount. The first image displaying the position of the image area obtained as a result of the first search process and the second image displaying the position of the image area obtained as a result of the second search process are simultaneously or switched. A display control unit that can be displayed, and a search process selection unit that accepts selection of either the first search process or the second search process after the first image or the second image is displayed. It is an image inspection apparatus including an inspection setting storage unit which stores the search process selected by the search process selection unit as an inspection setting.

According to this configuration, when the pattern area used for the search process is set on the registered image, at least two feature amounts, that is, the first feature amount and the second feature amount are extracted from the image in the pattern area. These feature quantities are stored in the feature quantity storage unit. The search processing unit executes the first search process using the first feature amount, detects an image region having a correlation value equal to or higher than a predetermined value with the first feature amount, and also detects a second feature amount. A second search process is executed using the image to detect an image region having a correlation value of a predetermined value or more with the second feature amount. The position of the image area obtained as a result of the first search process is displayed on the first image, and the position of the image area obtained as a result of the second search process is displayed on the second image. Since these first images and second images are displayed by the display control unit, the user can confirm the result of the first search process and the result of the second search process, and which search process is performed. A first image and a second image can be used as materials for selection or determination. Since the search process selected by the user is stored in the inspection setting storage unit as the inspection setting, the search process is executed by the selected search process during the operation of the image inspection device.

Since the first search process can be referred to as a search process by the first search algorithm and the second search process can be referred to as a search process by the second search algorithm, in the present invention, the search algorithm is selected. You can also do it.

In the second invention, the display control unit has the highest correlation image region having the highest correlation value among the plurality of image regions having the correlation values equal to or higher than the predetermined value, and the correlation value less than the highest correlation image region. It is possible that the image area having the correlation value of is displayed in a different display form.

According to this configuration, it is possible to easily distinguish between the highest correlation image region having the highest correlation value and the image region having a correlation value less than that. For example, if the highest correlation image area is the first candidate for the inspection target area, the image area having a correlation value less than that is the second candidate or less, and the extent to which the second candidate or less exists is visually expressed. can do.

Examples of different display modes are changing the color of the border surrounding the highest correlated image area and the color of the border surrounding the image area having a correlation value less than that, changing the line type of the border, and changing the border. The thickness can be changed. Further, it is also one of the different display forms to change the color to be colored in the highest correlation image area and the color to be colored in the image area having a correlation value less than that.

In the third aspect of the invention, the display control unit displays a detection frame surrounding an image area having a correlation value equal to or higher than the predetermined value, and the larger the size of the correlation value, the larger the detection frame. It can be configured.

According to this configuration, the magnitude of the correlation value can be visually expressed. For example, if the highest correlated image area is the first candidate of the inspection target area, the image area having a correlation value less than that is the second candidate or less, and the correlation value of the second candidate or less is the correlation value of the first candidate. It is possible to visually express how far or close they are.

The fourth invention can be configured so that the predetermined value can be adjusted.

According to this configuration, for example, if an image area other than the highest correlation image area is not displayed when the predetermined value is lowered, it is determined that the detection accuracy of the image area to be inspected is high and stable detection is possible. can do.

In the fifth aspect of the invention, the search processing selection unit detects whether or not the first image or the second image displayed by the display control unit is selected, and the first image is selected. If it is detected, the first search process is selected, and if it is detected that the second image is selected, the second search process may be selected. can.

According to this configuration, when the user selects the first image displayed by the display control unit, the first search process is automatically selected, and the user selects the second image displayed on the display control unit. Then, the second search process is automatically selected. As a result, the search process can be selected by the user's intuitive operation.

The selection of the first image may be an operation of selecting the first image itself, an operation of selecting a frame surrounding the first image, or a selection operation using various buttons or the like. The same applies to the selection of the second image.

In the sixth aspect of the invention, the search processing unit acquires the processing time of the first search processing and the processing time of the second search processing, and the display control unit is acquired by the search processing unit. It may be configured to display the processing time of the first search processing and the processing time of the second search processing.

That is, since the feature amount used is different between the first search process and the second search process, there may be a difference in the time (processing time) from the start to the completion of the process. In the present invention, since the processing time of the first search processing and the processing time of the second search processing can be displayed, the user can grasp each processing time and select the search processing. It can be used as a material for.

In a seventh aspect of the invention, the search processing unit continuously acquires a plurality of the input images and executes the first search process and the second search process on each of the input images. The display control unit may be configured to update the first image and the second image.

According to this configuration, it is possible to show the user the result of executing the first search process and the second search process on a plurality of different input images. This makes it possible to easily compare the stability of the first search process and the second search process.

The eighth invention includes an item input unit for inputting common setting items for the first search process and the second search process, and the search processing unit inputs common setting items input in the item input unit. It may be configured to be applied to the first search process and the second search process.

According to this configuration, if there is an item common to both search processes as a setting item of the first search process and the second search process, both searches can be input by inputting as a common setting item in the item input unit. It is automatically applied to the process. Therefore, the input operation of the setting item can be simplified.

In the ninth invention, the common setting item is an angle range in which the search process is performed.

According to a tenth aspect of the present invention, the search processing unit changes parameters related to image processing applied to the input image, repeatedly executes the search processing stored in the inspection setting storage unit, and obtains a plurality of detection results. It is configured to be acquired, and the display control unit may be configured to be able to display a plurality of detection results acquired by the search processing unit at the same time or by switching.

That is, even when an appropriate search process is selected, the result of the search process may be stable or unstable depending on the image processing parameters applied to the input image. In the present invention, the parameters related to the image processing applied to the input image can be changed, the search process can be repeatedly executed, and a plurality of detection results can be acquired and then displayed, so that the parameters of the image processing suitable for operation can be displayed. Can be set.

According to the eleventh invention, the parameter may include any one of the presence / absence of filtering, the type of filtering, and the strength of filtering.

That is, the presence / absence of filter processing applied to the input image, the type of filter processing, and the strength of the filter processing all have a large effect on the result of the search processing, and one or more of these are changed for the search. By repeatedly executing the process, it is possible to find a parameter that stabilizes the result of the search process.

In the twelfth invention, the parameter may include the degree of image reduction.

According to the present invention, the first image obtained as a result of the first search process using the first feature amount extracted from the image in the pattern region used for the search process and the second feature amount are used. Since the second image obtained as a result of the second search process can be displayed, even a user who does not know the characteristics of a plurality of search algorithms or the compatibility with the inspection target can use an appropriate search algorithm. It can be easily decided.

It is a schematic diagram which shows the operation time of the image inspection apparatus which concerns on embodiment of this invention. It is a block diagram of an image inspection apparatus. It is a figure which shows the example of the parameter of a wound inspection tool. It is a figure which shows the example of the preprocessing filter. It is a figure which shows the example of the parameter of real-time shading correction. It is a figure which shows the example of the screen which performed the pattern area setting and the search area setting. It is a figure which shows the example of the screen which performed the inspection area setting and the defect size setting. It is a figure which shows the example of the item input part for defect inspection processing. It is a figure which shows the example of the item input part for a search process. It is a figure which shows the display example of the search processing result image. It is a figure corresponding to FIG. 10 when the detection level is lowered. It is a figure which shows the example which displayed a plurality of search processing result images after parameter change. It is a figure which shows the example which selected and displayed one from the image shown in FIG. It is a figure which shows the display example of the defect inspection processing image. It is a figure corresponding to FIG. 14 when the detection level is lowered. It is a figure which shows the example which displayed a plurality of defect inspection result images after parameter change. It is a figure which shows the example which selected and displayed one from the image shown in FIG. It is a figure which shows the display example of the type of a preprocessing filter. It is a flowchart which shows the selection procedure of a search process. It is a figure which shows the case where the search process is selected on the selection interface. It is a flowchart which shows the selection procedure of a defect inspection process. It is a figure which shows the case which the defect inspection process is selected on the selection interface.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. It should be noted that the following description of the preferred embodiment is essentially merely an example and is not intended to limit the present invention, its application or its use.

FIG. 1 is a schematic view showing an operation time of the image inspection apparatus 1 according to the embodiment of the present invention. The image inspection device 1 is configured to inspect a defect of the work W by using an image of a work W (object to be inspected) moving in a certain direction, and specifically, light with respect to the work W. A lighting unit 2, an image pickup unit 3, a control unit 4, a monitor 5, a keyboard 6 and a mouse 7 are provided at least, but devices other than these can also be provided. For example, an external control device 8 including a programmable logic controller (PLC) or the like is connected to the control unit 4. The external control device 8 may form a part of the image inspection device 1. Further, the external control device 8 does not have to be a component of the image inspection device 1.

FIG. 1 shows a case where a plurality of work W are conveyed in the direction indicated by the white arrow in FIG. 1 in a state of being placed on the upper surface of the conveying belt conveyor B in a factory or the like. The work W is an object to be inspected. The external control device 8 is a device for sequence control of the transport belt conveyor B and the image inspection device 1, and a general-purpose PLC can be used.

In the description of this embodiment, the transport direction of the work W by the transport belt conveyor B (moving direction of the work W) is the Y direction, and the direction orthogonal to the Y direction in the plan view of the transport belt conveyor B is the X direction. The direction orthogonal to the X direction and the Y direction (the direction orthogonal to the upper surface of the conveyor belt B for transportation) is defined as the Z direction, but this is defined only for convenience of explanation.

The image inspection device 1 can be used for visual inspection of the work W, that is, for inspecting the presence or absence of defects such as scratches, stains, and dents on the surface of the work W, and is also called a visual inspection device. can. During its operation, the image inspection device 1 receives an inspection start trigger signal that defines the start timing of defect inspection from the external control device 8 via a signal line. The image inspection device 1 performs imaging, lighting, and the like of the work W based on the inspection start trigger signal, and obtains an inspection image after a predetermined process. After that, the inspection result is transmitted to the external control device 8 via the signal line. As described above, during the operation of the image inspection device 1, the inspection start trigger signal is repeatedly input and the inspection result is output between the image inspection device 1 and the external control device 8 via the signal line. As described above, the input of the inspection start trigger signal and the output of the inspection result may be performed via the signal line between the image inspection device 1 and the external control device 8, or other signals (not shown). It may be done via a line. For example, a sensor for detecting the arrival of the work W and the image inspection device 1 may be directly connected, and an inspection start trigger signal may be input from the sensor to the image inspection device 1.

In addition to being configured with dedicated hardware, the image inspection device 1 may be configured by installing software on a general-purpose device, for example, installing an image inspection program on a general-purpose or dedicated computer. In the following example, a case where the image inspection program is installed on a dedicated computer in which hardware such as a graphic board is specialized for image inspection processing will be described.

Further, the image inspection device 1 may be an image inspection device to which the principle of so-called deflectometry is applied. In that case, the inspection object is irradiated with the pattern light whose illuminance is periodically changed by the illumination unit 2, and the inspection object irradiated with the pattern light is imaged by the image pickup unit 3. The illumination unit 2 sequentially irradiates a plurality of pattern lights whose phases of the illuminance distribution are shifted, and each time the pattern light is irradiated, the image pickup unit 3 images an inspection object and inspects based on the obtained plurality of luminance images. Generates phase data showing the shape of the object. By generating an inspection image based on the phase data and using this inspection image, it is possible to perform a defect inspection related to the shape of the inspection object. The image inspection method to which the principle of deflationmetry is applied may be a conventionally known method, and therefore detailed description thereof will be omitted. The generation of the inspection image is not limited to the application of the principle of deflation metrics, and the inspection image may be generated by another method.

The image inspection device 1 has a mode for visually inspecting the work W conveyed one after another by the transport belt conveyor B, that is, an operation mode (Run mode) for actually determining the quality of the work W, and a search process used for the inspection. It is configured to be switchable to a setting mode (non-Run mode) for selecting and setting algorithms, defect inspection algorithms, various parameters, preprocessing, and the like. Switching between the operation mode and the setting mode can be performed, for example, by the mode switching means on the user interface displayed on the monitor 5. The user makes various settings and selections in the setting mode and makes adjustments before performing an appearance inspection of the work W conveyed one after another by the conveyor belt conveyor B in the operation mode. In the setting mode, various settings and selections can be made on the user interface displayed on the monitor 5.

Basically, default values are set for various parameters, and if the user determines that the default values are optimal as parameter values, there is no particular need to adjust the parameter values. However, in reality, due to differences in the surrounding lighting environment, the mounting position of the imaging unit 3, the posture deviation of the imaging unit 3, the focus adjustment, etc., the default values cannot be used to obtain the desired result. In some cases. Therefore, in the setting mode, the operation mode can be switched to the setting mode on the monitor 5 and the contents of various parameters can be edited.

In the operation mode, a plurality of work W flowing on the line of the transport belt conveyor B are imaged by the image pickup unit 3, and it is determined whether or not the work W has scratches, defects, or the like. If it is determined that the work W has scratches, defects, etc., the determination is NG. On the other hand, if it is determined that the work W has no scratches or defects, the determination is OK. In this way, the image inspection device 1 determines the quality of the appearance of the work W by using the image obtained by capturing the image of the work W.

(Structure of lighting unit 2)
The lighting unit 2 is composed of a lighting device that illuminates the work W, for example, surface lighting, coaxial epi-illumination that emphasizes gloss, low-angle lighting that emphasizes scratches and engraved edges, and black light illumination that illuminates black light. Lighting devices that perform various types of lighting, such as dome lighting that irradiates diffused light from all directions, can be used. The illumination unit 2 is connected to the control unit 4 via a signal line 100a, and can be installed away from the image pickup unit 3 and the control unit 4.

Further, when performing the deflect metering process, the illumination unit 2 may be a pattern light illumination device that irradiates the work W with pattern light having a periodic illuminance distribution. The pattern light illuminating device can be composed of, for example, a plurality of light emitting diodes, a liquid crystal panel, an organic EL panel, a digital micromirror device (DMD), or the like, and may be simply called a lighting unit. Although the liquid crystal panel, the organic EL panel, and the DMD are not shown, those having a conventionally known structure can be used.

When a plurality of light emitting diodes are used in the lighting unit 2 composed of the pattern light illuminating device, the plurality of light emitting diodes may be arranged in a dot matrix to generate patterned light having a periodic illuminance distribution by controlling the current value. can. In the case of a liquid crystal panel and an organic EL panel, by controlling each panel, the light emitted from each panel can be made into a pattern light having a periodic illuminance distribution. In the case of a digital micromirror device, it is possible to generate and irradiate a pattern light having a periodic illuminance distribution by controlling the built-in micromirror surface. The configuration of the illumination unit 2 is not limited to that described above, and equipment or devices capable of generating patterned light having a periodic illuminance distribution can also be used.

A lighting trigger signal transmitted from the lighting control unit 42 of the control unit 4 shown in FIG. 2 is input to the lighting unit 2 via the signal line 100a. The lighting unit 2 is configured to illuminate the work W each time it receives a lighting trigger signal.

The illuminance distribution of the illumination unit 2 can irradiate light with a uniform in-plane, and for example, the light emission state can be changed from a dark surface emission state to a bright surface emission state. Further, when the deflect metering process is performed, the pattern light irradiating the work W can be not only a sine waveform but also a pattern light such as a triangular wave.

(Structure of Imaging Unit 3)
The image pickup unit 3 shown in FIG. 1 has a light receiving element and a condensing optical system. The light receiving element is an image sensor composed of an image pickup element such as a CCD (charge-coupled device) or a CMOS (complementary metal oxide semiconductor) that converts the intensity of light obtained through a condensing optical system into an electric signal. The condensing system optical system is an optical system for condensing light incident from the outside, and typically has one or more optical lenses. A conventionally well-known autofocus mechanism can be provided in the condensing optical system.

The image pickup unit 3 is connected to the control unit 4 shown in FIG. 2 via a signal line 100b. In this embodiment, the case where the image pickup unit 3 and the illumination unit 2 are separate bodies will be described, but the image pickup unit 3 and the illumination unit 2 may be integrated.

The image pickup trigger signal transmitted from the image input unit 41 of the control unit 4 is input to the image pickup unit 3 via the signal line 100b. The imaging unit 3 is configured to perform imaging of the work W each time it receives an imaging trigger signal that defines the timing for capturing image data.

(Configuration of monitor 5)
The monitor 5 shown in FIG. 1 is made of, for example, an organic EL display, a liquid crystal display, or the like, and can also be called a display unit. The monitor 5 is connected to the display control unit 43 of the control unit 4 shown in FIG. 2, for example, an image captured by the image pickup unit 3, various images generated based on the image captured by the image pickup unit 3, and various settings. It is configured to be able to display an image, a defect inspection result of the work W, an interface for operation, an interface for various settings, setting values, and the like. The monitor 5 can also display a plurality of setting images, inspection images, and the like at one time. By using the monitor 5 as a touch operation panel type monitor, the monitor 5 can be provided with various information input functions.

The user can make various settings by visually recognizing the monitor 5 in the setting mode, and can confirm the operating state by visually recognizing the monitor 5 in the operation mode.

(Structure of keyboard 6 and mouse 7)
The keyboard 6 and the mouse 7 shown in FIG. 1 are conventionally known computer operation devices, and are connected to an operation input unit 44 (shown in FIG. 2) of the control unit 4. By operating the keyboard 6 or the mouse 7, various information can be input to the control unit 4, and images and the like displayed on the monitor 5 can be selected and buttons can be operated. The mouse 7 is an example of a pointing device, and other pointing devices can also be used. Further, instead of the keyboard 6 and the mouse 7, or in addition to the keyboard 6 and the mouse 7, a computer operation device such as a voice input device and a touch operation panel can be used.

(Configuration of control unit 4)
The control unit 4 is a unit for controlling each part of the image inspection device 1, and can be configured by a CPU, an MPU, a system LSI, a DSP, dedicated hardware, or the like. The control unit 4 is equipped with various functions as described later, and these may be realized by a logic circuit or may be realized by executing software. A part or all of the control unit 4 may be incorporated in the monitor 5, or a part of the control unit 4 may be incorporated in the illumination unit 2 or the image pickup unit 3.

As an example shown in FIG. 2, the control unit 4 includes an image input unit 41, a lighting control unit 42, a display control unit 43, an operation input unit 44, a main control unit 45, and a storage unit 46. It has a unit 47, a feature amount extraction unit 48 for defect inspection, an item input unit 49, an image processing unit 50, and a selection unit 51. Although each part of the control unit 4 is described separately as described above, the same part may be configured to execute a plurality of types of processing, or may be further subdivided and linked to each other for one processing. May be configured to run.

Each of the above hardware is connected so as to be capable of two-way communication or one-way communication via an electrical communication path (wiring) such as a bus.

The external control device 8 shown in FIG. 1 is communicably connected to the control unit 4 via a communication unit (not shown). The communication unit is installed on the production line to recognize the arrival timing of the work W, and when a trigger input is received from a sensor (photoelectric sensor, etc., not shown) connected to the external control device 8, the external control device It functions as an interface (I / F) for receiving an image pickup trigger signal from 8. The communication unit also functions as an interface (I / F) for receiving a control program or the like transferred from the outside.

The main control unit 45 shown in FIG. 2 performs numerical calculation and information processing based on various programs, and also controls each part of the hardware. The main control unit 45 stores a CPU 45a that functions as a central arithmetic processing unit, a work memory 45b such as a RAM that functions as a work area when the main control unit 45 executes various programs, a start-up program, an initialization program, and the like. It is provided with a program memory 45c such as a ROM, a flash ROM, or an EEPROM.

The lighting control unit 42 transmits a lighting trigger signal to the lighting unit 2 based on a command from the CPU 45 of the main control unit 45 or the external control device 8. The illumination trigger signal may include information about illuminance and pattern light.

The image input unit 41 can be configured from an ASIC (Application Specific Integrated Circuit) or the like that captures image data obtained by image pickup by the image pickup unit 3. The image input unit 41 may include a frame buffer for buffering image data. Specifically, when the image input unit 41 receives the image pickup command of the image pickup unit 3 from the CPU 45a, the image input unit 41 transmits an image data acquisition signal (image capture trigger signal) to the image pickup unit 3. Then, the image input unit 41 captures the image data obtained by the image pickup after the image pickup is performed by the image pickup unit 3. The captured image data is once buffered (cached). Image processing based on the principle of deflectometry can be performed in a processing unit (not shown).

Operation signals from the keyboard 6, mouse 7, and other operation devices are input to the operation input unit 44. The operation input unit 44 is a part that functions as an interface (I / F) for receiving operation signals from the keyboard 6, the mouse 7, and the like based on the operation of the user.

The display control unit 43 is composed of a display DSP or the like for displaying an image on the monitor 5. The display control unit 43 may include a video memory such as a VRAM that temporarily stores image data when displaying an image. Based on the display command (display command) sent from the CPU 45a, a control signal for displaying a predetermined image is transmitted to the monitor 5. For example, a control signal is sent to the monitor 5 in order to display image data captured by the image pickup unit 3 (image processed based on the principle of deflectometry), image data after image processing, various user interfaces, and the like. Send. The display control unit 43 also transmits a control signal for displaying the operation contents of the user using the keyboard 6 and the mouse 7 on the monitor 5. Further, a pointer or the like that can be operated by the mouse 7 can also be displayed on the monitor 5.

(Inspection tool)
The image inspection device 1 includes a plurality of measurement modules for performing visual inspection. Here, the measurement module that executes the visual inspection is called an inspection tool. There are various inspection tools, and the main image processing tools are edge position measurement tool, edge angle measurement tool, edge width measurement tool, edge pitch measurement tool, area measurement tool, blob measurement tool, pattern search measurement tool. , Scratch measurement tools, etc. Although detailed description is omitted, each inspection tool can be realized by using various conventional image processing methods.

The image inspection device 1 has a plurality of inspection tools, for example, which can be displayed on the monitor 5 in the form of a list, and at the same time, the user operates the pointer displayed on the monitor 5 with the mouse 7 or the like to use the pointer. You can now select the inspection tool you want.

The outline of each inspection tool will be described below, but the name of the inspection tool is determined only for the convenience of explanation, and any name may be used.

Edge position measurement tool: On the screen where the image of the work W is displayed, by setting a window for the inspection area where the edge position is to be detected, multiple scans can be performed in any direction within the set inspection area. Detects the edge of (the part that switches from light to dark or the part that switches from dark to light). From a plurality of detected edges, the designation of one edge is accepted, and the position of the edge that has received the designation is measured.

Edge angle measurement tool: Two segments are set in the inspection area where the setting is accepted, and the inclination angle of the work W from the edge detected in each segment is measured. The tilt angle can be positive, for example, clockwise.

Edge width measurement tool: Within the inspection area where the setting is accepted, scan in any direction to detect multiple edges and measure the width between the detected multiple edges.

Edge pitch measurement tool: Scans in any direction to detect multiple edges within the inspection area where the settings have been accepted. Measure the maximum / minimum value and average value of the distances (angles) between multiple detected edges.

Area measurement tool: The image of the work W captured by the image pickup unit 3 is binarized to measure the area of the white region or the black region. For example, the area of the white region or the black region is measured by accepting the designation of the white region or the black region as a parameter to be measured.

Blob measurement tool: The work W image captured by the imaging unit 3 is binarized, and the number, area, and center of gravity position as parameters for a set of pixels (blobs) with the same brightness value (255 or 0). Etc. are measured. The blob measurement tool detects the size (area) of the scratch as the same size as the actual defect, but tends to detect fine noise as well. In addition, the processing time tends to fluctuate depending on the detection state.

Pattern search measurement tool: An image pattern (model image) to be compared is stored in a storage device in advance, and a portion of the captured image of the work 8 that is similar to the image pattern stored in advance. By detecting, the position, tilt angle, and correlation value of the image pattern are measured.

Scratch inspection tool: Within the inspection area where the setting is accepted, a small area (segment) is moved to calculate the average density value of the pixel values, and it is determined that a scratch exists at a position where the density difference is equal to or greater than the threshold value. That is, the scratch inspection is not an area determination in a black-and-white image obtained by a simple binarization process, but an visual inspection algorithm that searches for defects such as scratches and stains by comparing with the shading level of the surroundings. The scratch inspection tool can also be called a tool that performs a scratch detection process.

The detection principle of scratch inspection will be described below. Here, a case where the X direction is specified as the scratch detection direction will be described as an example.
(1) First, the average concentration is measured while moving a small area (segment) of an arbitrary size within the inspection area by a predetermined amount. The amount to be moved can be determined based on the size of the segment, for example, 1/4 of the segment size.
(2) Next, the difference between the maximum concentration and the minimum concentration within a predetermined distance (here, 4 segments in the X direction) in the detection direction including the segment of interest is measured. This value is the "scratch level" of the segment of interest.
(3) When the scratch level exceeds a preset threshold value, the attention segment is counted as a scratch. This count value is a test result called "scratch amount".
After that, the attention segment is shifted by the amount of movement in the region, and the above steps (1) to (3) are repeated.

The above has described the case where the X direction is specified as the scratch detection direction. The scratch detection direction can also be specified in the Y direction or the XY direction. Here, a processing method when the scratch detection direction is specified in the two-dimensional direction of XY will be described. When the XY direction is specified, the difference between the maximum density and the minimum density within a total of 16 segments, 4 segments in each of the XY directions including the segment of interest, is measured. According to the scratch inspection tool, only defects to be detected can be detected without being affected by the shape of the work W and uneven lighting. In addition, since the scratch inspection tool processes by segmenting, it is highly resistant to fine noise, but the size (area) of the detected scratch tends to be larger than the actual defect. ..

In addition, an OCR recognition tool that recognizes character strings by cutting out character information in the inspection area and collating it with dictionary data, etc., while shifting the window (area) set on the image, edge at the position of each window. Trend edge tool that has the function of repeating the detection of, the shade tool that has the function of measuring the average of the shades in the set window, the deviation, etc. The user can select the necessary image processing tool according to the inspection content. It should be noted that these inspection tools are merely representative examples of typical functions and methods for realizing them. The image inspection device 1 may be configured to include an inspection tool corresponding to all image processing. Further, since each inspection tool is realized based on an individual inspection algorithm, it can also be called an inspection algorithm.

(Parameter of inspection tool)
FIG. 3 shows an example of the parameters of the scratch inspection tool. The display example shown in FIG. 3 is a user interface for the user to manually set the parameters of the scratch detection tool, and can be displayed on the monitor 5 as the scratch inspection tool parameter setting unit 61. Each value of the scratch inspection tool parameter setting unit 61 and the like can be changed by operating the keyboard 6 and the mouse 7.

The parameters of the scratch inspection tool include detection direction, setting for each direction, ON / OFF of high-speed mode, scratch level, setting direction, segment size, comparison interval setting, movement amount, comparison segment interval, gain, and the like. The scratch inspection tool parameter setting unit 61 shown in FIG. 3 has a detection direction setting field 61a, a direction-specific setting field 61b, a high-speed mode ON / OFF setting field 61c, a scratch level setting field 61d, a setting direction setting field 61e, and a segment size setting. A column 61f, a comparison interval setting column 61g, a movement amount setting column 61h, a comparison segment interval setting column 61i, and a gain setting column 61j are provided.

The detection direction is a parameter indicating in which direction the scratch is extended. Here, select any of XY, X, and Y from the drop-down list. For example, when the scratch is extended in the vertical direction, by selecting the X direction, the scratch can be detected by taking the difference in the horizontal direction.

The direction-specific setting is specified when it is desired to set the image processing parameter to a different value in the X direction or the Y direction.

The high-speed mode is a parameter for switching ON / OFF of high-speed processing. When the high-speed mode is turned on, the values that can be selected as image processing parameters are limited (for example, a multiple of 4), but the internal processing can be speeded up.

The scratch level is a parameter that defines the difference from the average density of the surrounding pixels judged as scratches. For example, if 10 is specified, pixels having a difference of less than 10 from the average density of the surrounding pixels are not judged as scratches and 10 The above pixels are determined as scratches.

The setting direction is linked to the direction-specific setting, and when the direction-specific setting is OFF, the following image processing parameters are set in common with the direction set in the detection direction (common to XY in the example of FIG. 3). When the direction-specific setting is ON, the following image processing parameters are individually set for each direction.

The segment size is a parameter defined according to the size of the scratch. For example, in the example of FIG. 3, 8 is specified as the segment size, an 8 × 8 area is taken around a part having a shape image to be inspected, this is divided into 4 parts, and the difference for each area is taken to obtain a scratch. Determine the amount. In this way, the segment size is a parameter that defines the size of the area where scratches are detected.

Select manual or automatic for the comparison interval setting.

The movement amount is moved every pixel (here, 4 pixels) for which the comparison interval is specified.

The comparison segment interval defines the interval between the segments that compare the concentration differences.

(Pretreatment filter)
The image inspection device 1 includes a plurality of image processes applied to the inspection image as a preprocessing filter. FIG. 4 shows an example of a preprocessing filter. The example shown in FIG. 4 is a user interface for manually setting a preprocessing filter that the user wants to apply to the inspection image, and can be displayed on the monitor 5 as the preprocessing filter setting unit 62. The preprocessing filter setting unit 62 is provided with a drop-down list button 62a, and when the drop-down list button 62a is clicked with the mouse 7, a preprocessing filter as shown in FIG. 4 is displayed, and the mouse 7 displays the preprocessing filter. You can choose. If you do not need the preprocessing filter, you can select "None". The preprocessing filter includes a "real-time shading correction" (shading correction) filter. The shading correction has a light / dark selection function such as extracting only a bright part or a dark part with respect to the periphery.

FIG. 5 shows an example of parameters for real-time shading correction. The example shown in FIG. 5 is a user interface for the user to manually set the parameters of the real-time shading correction, and can be displayed on the monitor 5 as the real-time shading correction parameter setting unit 63. Each value of the real-time shading correction parameter setting unit 63 and the like can be changed by operating the keyboard 6 and the mouse 7.

Parameters of real-time shading correction include correction method, extraction size, processing direction, extraction color, gain, noise removal, contrast equalization, out-of-area reference, number of times, and the like. The real-time shading correction parameter setting unit 63 includes a correction method setting field 63a, an extraction size setting field 63b, a processing direction setting field 63c, an extraction color setting field 63d, a gain setting field 63e, a noise removal setting field 63f, and a contrast equalization setting field 63g. , The out-of-area reference setting field 63h and the number of times setting field 63i are provided.

(Structure of storage unit 46)
The storage unit 46 shown in FIG. 2 can be configured by, for example, a RAM or the like, and includes a registered image storage unit 46a, a feature amount storage unit 46b for search processing, and an inspection setting storage unit 46c. The registered image storage unit 46a is a portion that stores the registered image obtained by imaging the work W. The registered image can be an image captured by the image pickup unit 3 on the work W in the setting mode. The registered image is used when setting a pattern area used for the search process described later.

The search processing feature amount storage unit 46b is a portion that stores the first search processing feature amount extracted from the image in the pattern region and the second search processing feature amount. The first feature amount for search processing is, for example, a shade image that can be extracted from an image in a pattern region, and can be a normalized correlation amount. The second search processing feature amount is edge information (geometric information such as edge image, orientation, and size) that can be extracted from the image in the pattern area.

The inspection setting storage unit 46c stores the search process (search algorithm) selected by the search process selection unit described later as an inspection setting, and inspects the defect inspection process (defect inspection algorithm) selected by the inspection process selection unit 51b. This is the part to be stored as a setting. The selected search process and the selected defect inspection process are stored in the inspection setting storage unit 46c in the setting mode. In the operation mode, the search process and the defect inspection process stored in the inspection setting storage unit 46c are read from the inspection setting storage unit 46c and applied.

(Structure of setting unit 47)
The setting unit 47 includes a pattern area setting unit 47a, an inspection area setting unit 47b, and a defect size setting unit 47c. The pattern area setting unit 47a is a part for reading the registered image stored in the registered image storage unit 46a and setting a pattern area to be used for the search process on the registered image.

Specifically, the pattern area setting unit 47a controls the display control unit 43 after reading the registered image stored in the registered image storage unit 46a, and monitors the registered image 64 as shown in FIG. To display. The pattern area setting unit 47a detects that the user has drawn a frame line 64a so as to surround the pattern area by operating the mouse 7. By moving the mouse 7 while the user clicks the button of the mouse 7, for example, the frame line 64a surrounding the pattern area can be drawn on the registered image 64. Further, by clicking the edge of the pattern area at a plurality of places, the frame line 64a can be automatically drawn so as to surround the pattern area. The pattern area setting unit 47a detects the operation of the mouse 7 as described above and accepts the setting of the pattern area used for the search process on the registered image 64. The pattern area can also be stored in the storage unit 46 of the control unit 4.

Further, the setting unit 47 shown in FIG. 2 is configured to be able to accept the setting of the search area. The setting of the search area can be accepted on the registered image 64, and the frame line 64b is drawn so as to surround the search area by the operation of the mouse 7 in the same manner as the setting of the pattern area. The setting unit 47 detects the operation of the mouse 7 by the user as described above, and accepts the setting of the search area used for the search process. The search area can also be stored in the storage unit 46 of the control unit 4. Since the search area is an area for performing search processing and does not perform search processing for areas other than the search area, the smaller the search area, the shorter the time required for search processing.

The inspection area setting unit 47b is a part that accepts the setting of the inspection area on the inspection image. The inspection area is an area for performing defect inspection processing, and defect inspection processing is not performed for areas other than the inspection area. Therefore, the smaller the inspection area, the shorter the time required for defect inspection processing. The inspection image can be an image captured by the image pickup unit 3 in the setting mode, and can be stored in the storage unit 46 in the same manner as the registered image. Therefore, the inspection area setting unit 47b is a part for reading the inspection image stored in the storage unit 46 and setting the defect inspection process on the inspection image.

Specifically, the inspection area setting unit 47b controls the display control unit 43 after reading the inspection image stored in the storage unit 46, and displays the inspection image 65 on the monitor 5 as shown in FIG. Let me. Similar to the setting of the pattern area, the user draws a frame line 65a so as to surround the inspection area by operating the mouse 7. The inspection area setting unit 47b detects the operation of the mouse 7 by the user as described above, and receives the setting of the inspection area used for the defect inspection process on the inspection image 65. The inspection area can also be stored in the storage unit 46 of the control unit 4.

The defect size setting unit 47c shown in FIG. 2 is a portion that receives from the user the setting of the defect size, which is the size of the image area to be detected as a defect in the inspection area received by the inspection area setting unit 47b. The defect size is smaller than the size of the inspection area. Specifically, when the defect portion is indicated by reference numeral 65c in FIG. 7, the minimum frame line (rectangular region) 65b in which all the defect portions 65c can be inserted is provided. The defect size can be set by drawing by operating the mouse 7. The defect size setting unit 47c can detect the operation of the mouse 7 as described above and accept the defect size setting on the inspection image 65. In this way, since the defect size is set by the user, it is an accurate one that reflects the actual size of the defect.

The defect size setting unit 47c can be configured to enable at least one input of, for example, a numerical value of the defect size and an input by drawing the defect portion, in addition to the input by drawing the rectangular area described above. .. As a method of inputting the numerical value of the defect size, for example, a method of inputting the vertical dimension and the horizontal dimension of the defect portion 65c by operating the keyboard 6 or the mouse 7, or the number of vertical pixels and the horizontal pixel of the defect portion 65c. There is a method to enter the number. As an input method by drawing the defective portion, for example, a method of operating the mouse 7 so that the pointer of the mouse 7 traces the defective portion 65c on the inspection image 65, or a method of tracing the defective portion 65c with a pressure-sensitive pen. There are methods and so on. Regardless of any of these input methods, the defect size setting unit 47c accepts the defect size setting.

(Structure of feature amount extraction unit 48 for defect inspection)
The defect inspection feature amount extraction unit 48 shown in FIG. 2 has the first defect inspection feature amount and the first defect inspection feature amount for each pixel in the inspection area 65a (shown in FIG. 7) received by the inspection area setting unit 47b. This is a part for extracting a second defect inspection feature amount different from the defect inspection feature amount of the above. The first defect inspection feature amount is the difference in shading value from the peripheral pixels, and the second defect inspection feature amount is the shading value. Therefore, the defect inspection feature amount extraction unit 48 obtains the difference in the shading value from the peripheral pixels for each pixel in the inspection region 65a, and also obtains the shading value. Based on these, the shading value with the peripheral pixels is obtained. The difference is extracted as the first defect inspection feature amount, and the shading value is extracted as the second defect inspection feature amount. The extracted first defect inspection feature amount and second defect inspection feature amount can be stored in the storage unit 46.

(Structure of item input unit 49)
The item input unit 49 shown in FIG. 2 is a part for inputting common setting items for the first defect inspection process and the second defect inspection process. When the first defect inspection process is a scratch inspection tool and the second defect inspection process is a blob measurement tool, the common setting item can be the light / dark designation of the defect. Specifying the brightness of a defect can also be said to be the input of brightness information.

Specifically, as shown in FIG. 8, the item input unit 49 controls the display control unit 43 to display the inspection image 65 on the monitor 5, and is set in common to the monitor 5 on which the inspection image 65 is displayed. The item input field 66 is displayed. A drop-down list button 66a is provided in the common setting item input field 66, and when the drop-down list button 66a is clicked with the mouse 7, as shown in FIG. 8, "black and white" is selected as a light / dark selection option. Three options are displayed: "Both", "White scratches only", and "Black scratches only". One of the three selection branches can be selected by the keyboard 6 or the mouse 7, and the item input unit 49 detects this selection operation and accepts the input of the common setting item. The received common items can be stored in the storage unit 46, and are applied to the first defect inspection process and the second defect inspection process.

The item input unit 49 is also a part capable of inputting common setting items for the first search process and the second search process. When the first search process is a pattern search and the second search process is an edge-based search, the common setting item can be an angle range in which the search process is performed. If there is a possibility that the set pattern area may appear in an inclined state during the search process of the operation mode, set the angle range. The angle range can be set, for example, as "+" in the clockwise direction and "-" in the counterclockwise direction.

Specifically, as shown in FIG. 9, the item input unit 49 controls the display control unit 43 to display the registered image 64 on the monitor 5, and is set in common to the monitor 5 on which the registered image 64 is displayed. The item input field 67 is displayed. An angle range input unit 67a is provided in the common setting item input field 67, and when the angle range input unit 67a is clicked with the mouse 7, the angle can be input numerically. It should be noted that the numerical value may be directly input by operating the keyboard 6.

The item input unit 49 detects the input operation of the keyboard 6 or the mouse 7 and accepts the input of the common setting item. As shown in FIG. 9, when "30" is input, the image area is detected by the search process even if the center line of the set pattern area is tilted by plus or minus 30 ° with respect to the vertical direction of the monitor 5. Will be possible. The common items received by the item input unit 49 can be stored in the storage unit 46, and are applied to the first search process and the second search process.

(Image processing unit 50)
The image processing unit 50 shown in FIG. 2 includes a search processing unit 50a and a defect inspection unit 50b.

(Structure of selection unit 51)
The selection unit 51 shown in FIG. 2 includes a search processing selection unit 51a and an inspection processing selection unit 51b.

(Structure of search processing unit 50a)
In addition to the registered image 64, an image obtained by newly imaging the work W by the imaging unit 3 is input to the search processing unit 50a. An image obtained by newly imaging the work W by the imaging unit 3 is called an input image. The search processing unit 50a has a first search process for detecting an image region having a correlation value (score) equal to or higher than a predetermined value with a feature amount for the first search process and a second search process for the input image. A second search process for detecting an image region having a correlation value between the feature amount and a predetermined value or more is executed. The first search processing feature amount and the second search processing feature amount used by the search processing unit 50a are read from the search processing feature amount storage unit 46b.

When the feature amount for the first search processing is the normalized correlation amount, the first search processing can be a pattern search. The pattern search is a normalized correlation search, in which a correlation value (similarity) between a pattern area and a part of an input image is calculated, and an image area having a high correlation value (predetermined value or more) is calculated based on the correlation value. To detect.

When the feature amount for the second search processing is edge information, the second search processing can be an edge-based search. The edge-based search obtains the edge information of the pattern area and the edge information of a part of the input image, calculates the correlation value (similarity) of both edges, and has a high correlation value based on the correlation value (the correlation value is high). (More than a predetermined value) Detects an image area.

The search processing unit 50a acquires the processing time of the first search processing and the processing time of the second search processing. That is, the time from the start to the end of the first search process and the time from the start to the end of the second search process are measured and stored in the storage unit 46. be able to. The first search process and the second search process may be performed in parallel, or may be performed at different timings.

The display control unit 43 shown in FIG. 2 determines the position of the first search processing result image displaying the position of the image area obtained as a result of the first search processing and the position of the image area obtained as a result of the second search processing. It is configured to simultaneously display the displayed second search processing result image on the monitor 5. As shown in FIG. 10, the display control unit 43 causes the monitor 5 to display the search result image display user interface 68. The user interface 68 for displaying the search result image has a first image display area 68a in which the first search processing result image 71 displaying the position of the image area obtained as a result of the first search processing is displayed, and a second image display area 68a. A second image display area 68b on which the second search processing result image 72 displaying the position of the image area obtained as a result of the search processing is displayed, and a detection level adjustment area 68c are provided. The first search processing result image 71 and the second search processing result image 72 are images in which the position of the image region is superimposed and displayed on the input image.

In this embodiment, the first search processing result image 71 and the second search processing result image 72 are simultaneously displayed on the monitor 5, but the present invention is not limited to this, and the first search processing result image 71 and the second search processing are not limited to this. The result image 72 may be switched and displayed on the monitor 5. For example, the search result image display user interface 68 is provided with an image switching means (eg, a button or the like), and the user operates the image switching means with a mouse 7 or the like to obtain only the first search processing result image 71. Can be displayed, or only the second search processing result image 72 can be displayed. Further, for example, tabs may be provided around each of the first search processing result image 71 and the second search processing result image 72, and by operating the tabs, only the first search processing result image 71 may be displayed. , It is also possible to display only the second search processing result image 72.

The display control unit 43 includes an image region having the highest correlation value among a plurality of image regions having a correlation value of a predetermined value or more, and an image region having a correlation value less than the correlation value of the highest correlation image region. Is configured to be displayed in a different display format. For example, in the first image 71 shown in FIG. 10, the largest frame line (detection frame) 71a and the frame line (detection frame) 71b smaller than the frame line 71a are superimposed and displayed on the input image. The region surrounded by the frame line 71a is the highest correlation image region, and this region is the image region having the highest correlation value. On the other hand, the region surrounded by the frame line 71b is an image region having a lower correlation value than the region surrounded by the frame line 71a. The size of the frame line 71a and the frame line 71b is related to the size of the correlation value, and the display control unit 43 is configured so that the larger the size of the correlation value, the larger the frame line. Similar frame lines 72a and 72b (shown in FIG. 11) can be displayed on the second search processing result image 72 as well. The arrows in the frame lines 71a and 71b indicate the posture of the detected image area.

In the pattern search, as shown in the first search processing result image 71, a plurality of image areas are displayed, and the pattern opposite to the pattern to be detected is detected with a high correlation value. It can be judged that the stability is low. On the other hand, in the edge-based search, as shown in the second search processing result image 72, the detected image area is one and the correlation value is high, so that the image area can be stably detected. It can be judged that.

Therefore, the user can detect the image area with a correlation value by looking at the first search processing result image 71 and the second search processing result image 72 displayed on the search result image display user interface 68. You can easily grasp whether it is.

The first image display area 68a of the search result image display user interface 68 is provided with a first processing time display area 68d for displaying the processing time of the first search processing acquired by the search processing unit 50a. Further, the second image display area 68b of the search result image display user interface 68 is provided with a second processing time display area 68f for displaying the processing time of the second search processing acquired by the search processing unit 50a. .. Thereby, the user can easily compare the processing time of the first search processing and the processing time of the second search processing. The correlation value can be displayed in the range of 0 to 100.

The first image display area 68a of the search result image display user interface 68 is provided with a first correlation value display area 68e for displaying the correlation value of the highest correlation image area. Further, the second image display area 68b of the search result image display user interface 68 is provided with a second correlation value display area 68g for displaying the correlation value of the highest correlation image area. Thereby, the user can easily compare the highest correlation value of the first search process and the highest correlation value of the second search process.

The detection level adjustment area 68c is an area for adjusting the detection level (predetermined value). The detection level can be a correlation value. For example, by inputting "60" in the detection level adjustment area 68c, only the image area having the correlation value of 60 or more is extracted and the first image display area 68a and the second image are used. It can be displayed in the display area 68b. The detection level input to the detection level adjustment area 68c is applied to both the first search process and the second search process. That is, the detection level used in the first search process and the detection level used in the second search process can be collectively adjusted. The detection level used in the first search process and the detection level used in the second search process are the same values.

For example, as shown in FIG. 11, when "40" is input to the detection level adjustment area 68c to lower the detection level, the display control unit 43 performs the first search processing result image 71 and the second search after the detection level adjustment. The processing result image 72 is displayed on the monitor 5. The first search processing result image 71 and the second search processing result image 72 after the detection level adjustment may be displayed at the same time or may be switched and displayed.

When the detection level is lowered, the number of image areas displayed in the first image display area 68a and the second image display area 68b increases. If the number of image regions does not increase even if the detection level is lowered, it means that the difference between the highest correlation image region and the image region having a correlation value less than that is large, and the detection of the image region is stable. It can be said that it is doing. Therefore, the stability of detection of the image region can be grasped by adjusting the detection level.

In this embodiment, in order to easily distinguish between the highest correlated image region and the image region having a correlation value less than that, these image regions are displayed in different display modes. As an example, a form in which the sizes of the frame lines 71a and 71b are changed according to the correlation value is shown, but as an example of different display forms, the color of the frame line surrounding the highest correlation image area and the correlation value less than that are provided. The color of the border surrounding the image area can be changed, the line type of the border can be changed, the thickness of the border can be changed, and the like. Further, it is also one of the different display forms to change the color to be colored in the highest correlation image area and the color to be colored in the image area having a correlation value less than that. Further, the correlation value may be displayed in the highest correlation image area and the image area having a correlation value less than that.

The search processing unit 50a applies the common items received by the item input unit 49 to the first search processing and the second search processing. As a result, it is not necessary to input items to be applied to the first search process and the second search process a plurality of times, for example, an angle range.

The search processing unit 50a continuously acquires a plurality of input images and executes a first search process and a second search process for each input image. The display control unit 43 is configured to update the first search processing result image 71 and the second search processing result image 72. It may be difficult to estimate the stability of the search process just by looking at the result of one search process, but the first search process is executed for each of the different input images, and a plurality of first search processes are performed. 1 Search processing result The stability of the first search processing can be estimated by obtaining the image 71 and sequentially displaying it on the monitor 5. The same applies to the second search process. It is also possible to display the plurality of first search processing result images 71 and the second search processing result images 72 on the monitor 5 to check the stability.

(Structure of search processing selection unit 51a)
After the first search processing result image 71 and the second search processing result image 72 are displayed on the search result image display user interface 68 shown in FIG. 11, the search processing selection unit 51a may perform the first search processing or the second search processing. This is the part that accepts the selection of any of the search processes. Specifically, the search processing selection unit 51a detects whether or not the first search processing result image 71 or the second search processing result image 72 displayed by the display control unit 43 is selected, and the first search processing result. When it is detected that the image 71 is selected, the first search process is selected, and when it is detected that the second search process result image 72 is selected, the second search process is selected. It is configured.

That is, when the user selects the first search processing result image 71 displayed on the monitor 5, the first search processing is automatically selected, and the user selects the second search processing result image 72 displayed on the monitor 5. When is selected, the second search process is automatically selected. As a result, the search process can be selected by the user's intuitive operation.

The selection of the first search processing result image 71 may be an operation of selecting the first search processing result image 71 itself, an operation of selecting a frame surrounding the first search processing result image 71, and various buttons. It may be a selection operation by such as. The same applies to the selection of the second search processing result image 72. Such an operation can be detected, for example, by detecting the pointer position of the mouse 7 and detecting the click operation. Alternatively, one may be selected from the tool selection drop-down list in the detection level adjustment area 68c of the search result image display user interface 68 shown in FIG. When the "Next" button in the interface of the detection level adjustment area 68c is pressed after selecting the tool, the tool used for the inspection is confirmed.

(Parameter change in search process)
The search processing unit 50a changes the parameters (search processing parameters) related to the image processing applied to the input image, repeatedly executes the search processing stored in the inspection setting storage unit 46c, and acquires a plurality of detection results. It is configured to do. The display control unit 43 is configured to be able to display a plurality of detection results acquired by the search processing unit 50a at the same time or by switching.

The search processing parameter includes any one or more of the presence / absence of filter processing, the type of filter processing, and the intensity of filter processing, and in particular, the degree of image reduction can be included. The search processing parameters may include all of the presence / absence of filtering processing, the type of filtering processing, and the strength of filtering processing. The items and values included in the search processing parameters may differ between the first search processing and the second search processing.

In the case of edge-based search, the parameters for search processing can include parameters that are frequently adjusted in edge-based search, for example, "image reduction" for edge extraction and "edge extraction threshold". There are two parameters.

In FIG. 12, when the tool determined by the search processing selection unit 51a is an edge-based search, the parameters related to the image processing applied to the input image are changed and the search processing is performed, and the images obtained as a result are displayed in a list. This is an example of the above, and the list display interface 90 is shown. In this example, since the tool determined by the search process selection unit 51a is the second search process (edge-based search), the image shown in FIG. 12 is the second search process result image 72, but the search process is performed. When the tool determined by the selection unit 51a is the first search process (pattern search), the first search process result image 71 can be displayed as a list display form as shown in FIG. Hereinafter, the case where the second search processing result image 72 is displayed will be described, but the same applies to the case where the first search processing result image 71 is displayed.

In the list display interface 90 shown in FIG. 12, the images arranged in the vertical direction are images in which the "image reduction degree" is changed, and the images arranged in the horizontal direction are images in which the "edge extraction threshold value" is changed. .. Since eight parameters are set, eight second search processing result images 72 are displayed. However, the present invention is not limited to this, and two or more parameters may be set. Therefore, the list display interface 90 is displayed. Can display at least two second search processing result images 72. In the list display interface 90, the displayed image can be switched or enlarged and displayed.

The list display interface 90 is provided with a detection level adjustment area 90a. The detection level adjustment area 90a is the same as the detection level adjustment area 68c shown in FIGS. 10 and 11, and the detection level can be adjusted while the second search processing result image 72 is displayed in a list. The adjusted detection level is applied to all the second search processing result images 72.

The user can select any one second search processing result image 72 from the plurality of second search processing result images 72 displayed on the list display interface 90. The selection operation of the second search processing result image 72 by the user is accepted by, for example, the operation input unit 44. For example, when the user puts the pointer of the mouse 7 on the image displayed as "7" from the plurality of second search processing result images 72 displayed on the monitor 5 and clicks the image, the selection of the image is completed. do.

The selection of the second search processing result image 72 may be an operation of selecting the second search processing result image 72 itself, an operation of selecting a frame surrounding the second search processing result image 72, and various buttons. It may be a selection operation by such as. Such an operation can be detected, for example, by detecting the pointer position of the mouse 7 and detecting the click operation. Alternatively, "7" may be selected from the image selection drop-down list in the detection level adjustment area 90a of the list display interface 90 shown in FIG. When the "End" button in the interface of the detection level adjustment area 90a is pressed after the image is selected, the parameters used for the inspection are confirmed.

For example, instead of the "End" button in the detection level adjustment area 90a of the list display interface 90 shown in FIG. 12, the "Next" button is used. By making the parameter changes multi-layered, such as adjusting the parameters of the second stage related to the changed parameters, it may be possible to set the parameters in detail. In the case of multi-layering, the image selection screen as shown in FIG. 13 appears a plurality of times in the process of image selection.

It can also be configured to accept an operation to return to the previous selection screen by pressing the "forward" button in the detection level adjustment area 90a. By returning to the previous screen while maintaining the parameters selected in the image, it is possible to return to the screen shown in FIG. 10 in a state adjusted to the preferable parameters and compare the inspection tools with each other.

FIG. 13 shows a selection display interface 91 that enlarges and displays the selected second search processing result image 72 for final confirmation after the parameters are determined. The selection display interface 91 is provided with a parameter adjustment area 91a. The parameter adjustment area 91a is provided with an operation button or the like for individually changing the parameters, and the user can change or finely change the parameters as necessary even after selecting the second search processing result image 72. Adjustment is possible.

Image processing that enhances the stability of the search processing by combining with the search processing selected by the search processing selection unit 51a is set in advance, and the image processing is automatically performed for the input image to be selected for the search processing. It can also be configured as follows. For example, the normalization correlation search process is compatible with a smoothing filter in a large range such as "blurring process" as a preprocessing filter in order to cancel the influence of fine shading changes of the work W, so it is normalized. In the correlation search process, "blurring process" is automatically applied to the input image. In the case of the normalization correlation search processing, it is preferable to obtain a plurality of search processing result images as shown in FIG. 12 by changing the intensity of the “blurring process” and the “image reduction degree”.

(Structure of defect inspection unit 50b)
The defect inspection unit 50b is a portion that executes the first defect inspection process and the second defect inspection process. In the first defect inspection process, whether or not a set of pixels having a first defect inspection feature amount larger than a predetermined threshold value should be detected as a defect based on the defect size set by the defect size setting unit 47c. Is a process for determining. Whether or not the second defect inspection process should detect a set of pixels having a second defect inspection feature amount larger than a predetermined threshold value as a defect based on the defect size set by the defect size setting unit 47c. Is a process for determining. The defect size used in the first defect inspection process is the same as the defect size used in the second defect inspection process. The first defect inspection process and the second defect inspection process are executed by using the input image obtained by newly imaging the work W by the image pickup unit 3. Further, the first defect inspection feature amount and the second defect inspection feature amount are extracted by the defect inspection feature amount extraction unit 48.

When the first defect inspection feature amount is the difference in shading value from the peripheral pixels, the first defect inspection process can be a scratch inspection tool. When the feature amount for the second defect inspection is a shading value, the second defect inspection process can be a blob measurement tool.

The defect inspection unit 50b acquires the processing time of the first defect inspection process and the processing time of the second defect inspection process. That is, the time from the start to the end of the first defect inspection process and the time from the start to the end of the second defect inspection process are measured and stored in the holding or storage unit 46, respectively. Can be kept. The first defect inspection process and the second defect inspection process may be performed in parallel or may be performed at different timings.

The display control unit 43 displays a first defect inspection processed image displaying the defect area detected by the first defect inspection process and a second defect inspection processed image displaying the defect area detected by the second defect inspection process. Is configured to be displayed on the monitor 5 at the same time.

As shown in FIG. 14, the display control unit 43 causes the monitor 5 to display the defect inspection result image display user interface 78. The defect inspection result image display user interface 78 includes a first image display area 78a on which a first defect inspection processing image 81 displaying the position of a set of pixels obtained as a result of the first defect inspection processing is displayed. A second image display area 78b on which a second defect inspection processed image 82 displaying the position of a set of pixels obtained as a result of the second defect inspection process is displayed, and a detection level adjustment area 78c are provided. The first defect inspection processed image 81 and the second defect inspection processed image 82 are images in which the positions of a set of pixels are superimposed and displayed on the input image.

In this embodiment, the first defect inspection processed image 81 and the second defect inspection processed image 82 are simultaneously displayed on the monitor 5, but the present invention is not limited to this, and the first defect inspection processed image 81 and the second defect inspection are not limited to this. The processed image 82 may be switched and displayed on the monitor 5. For example, the defect inspection result image display user interface 78 is provided with an image switching means (eg, a button or the like), and the user operates the image switching means with a mouse 7 or the like to obtain a first defect inspection processed image 81. Only the second defect inspection processed image 82 can be displayed. Further, for example, tabs may be provided around each of the first defect inspection processed image 81 and the second defect inspection processed image 82, and by operating the tabs, only the first defect inspection processed image 81 may be displayed. , Only the second defect inspection processed image 82 can be displayed.

The display control unit 43 is configured to attach a mark 81a to the defect area detected by the first defect inspection process and display the mark 81a superimposed on the first defect inspection process image 81. Examples of the mark 81a include a + mark, an arrow, a circle mark, and the like. Further, the frame line surrounding the defect region may be configured to be superimposed and displayed on the first defect inspection processed image 81.

Further, the display control unit 43 attaches a mark 82a similar to that in the case of the first defect inspection processed image 81 to the defect region detected by the second defect inspection process, and uses this mark 82a as the second defect inspection processed image. It is configured to be superimposed and displayed on 82. Further, the frame line surrounding the defect region may be configured to be superimposed and displayed on the second defect inspection processed image 82.

The first image display area 78a of the defect inspection result image display user interface 78 is provided with a first processing time display area 78d for displaying the processing time of the first defect inspection process acquired by the defect inspection unit 50b. .. Further, the second image display area 78b of the defect inspection result image display user interface 78 is provided with a second processing time display area 78f for displaying the processing time of the second defect inspection process acquired by the defect inspection unit 50b. ing. This allows the user to easily compare the processing time of the first defect inspection process with the processing time of the second defect inspection process.

Comparing the first defect inspection processed image 81 obtained by the scratch inspection tool and the second defect inspection processed image 82 obtained by the blob measurement tool, the first defect inspection processed image 81 is segmented by the scratch inspection tool. It can be seen that the size (area) of the detected scratches tends to be larger than the actual defects, although the resistance to fine noise is strong because the processing is performed. On the other hand, in the second defect inspection processed image 82 obtained by the blob measurement tool, the detected scratch size (area) is detected to be the same size as the actual defect, but fine noise is also detected. It tends to be easy. Further, in the case of the blob measurement tool, the processing time tends to fluctuate depending on the detection state.

The detection level adjustment area 78c is an area for adjusting the detection level (threshold value). The detection level can be a gradation value. For example, by inputting "50" in the detection level adjustment area 78c, a set of pixels having a gradation value larger than 50 is extracted and the first image display area 78a and It can be displayed in the second image display area 78b. The detection level input to the detection level adjustment region 78c is applied to both the first defect inspection process and the second defect inspection process. That is, the detection level used in the first defect inspection process and the detection level used in the second defect inspection process can be collectively adjusted. The detection level used in the first defect inspection process and the detection level used in the second defect inspection process are the same values. In the first defect inspection process using the scratch inspection tool, a portion where the amount of change from the periphery of the average concentration value of the small region (segment) exceeds the specified detection level is detected. In the second defect inspection process using the blob measurement tool, a portion having a density value exceeding the detection level is detected with respect to the average value of the shading values of the entire screen.

For example, as shown in FIG. 15, when "30" is input to the detection level adjustment area 78c to lower the detection level, the display control unit 43 displays the first defect inspection processed image 81 and the second defect after the detection level adjustment. The inspection processing image 82 is displayed on the monitor 5. The first defect inspection processed image 81 and the second defect inspection processed image 82 after the detection level adjustment may be displayed at the same time or may be switched and displayed.

When the detection level is lowered, the number of sets of pixels displayed in the first image display area 78a and the second image display area 78b increases. By changing the detection level, the defect detection state can be adjusted, and the stability of the detection process due to the difference in the detection level can be compared.

The defect inspection unit 50b applies the common items received by the item input unit 49 to the first defect inspection process and the second defect inspection process. As a result, it is not necessary to input the items applied to the first defect inspection process and the second defect inspection process a plurality of times, for example, as in the case of light / dark designation.

The defect inspection unit 50b continuously acquires a plurality of inspection images and executes a first defect inspection process and a second defect inspection process on each inspection image. The display control unit 43 is configured to update the first defect inspection processed image 81 and the second defect inspection processed image 82. It may be difficult to estimate the stability of the defect inspection process just by looking at the result of one defect inspection process, but the first defect inspection process is executed for each of the different input images. By obtaining a plurality of first defect inspection processed images 81 and sequentially displaying them on the monitor 5, the stability of the first defect inspection process can be estimated. The same applies to the second defect inspection process. It is also possible to display the plurality of first defect inspection processed images 81 and the second defect inspection processed image 82 on the monitor 5 to check the stability.

(Structure of inspection process selection unit 51b)
After the first defect inspection process image 81 and the second defect inspection process image 82 are displayed on the defect inspection result image display user interface 78 shown in FIG. 14, the inspection process selection unit 51b performs the first defect inspection process or the first defect inspection process. It is a part that accepts the selection of any one of the defect inspection processes of 2. Specifically, the inspection processing selection unit 51b detects whether or not the first defect inspection processing image 81 or the second defect inspection processing image 82 displayed by the display control unit 43 is selected, and the first defect inspection processing When it is detected that the image 81 is selected, the first defect inspection process is selected, and when it is detected that the image 82 is selected, the second defect inspection process is selected. It is configured as follows.

That is, when the user selects the first defect inspection processed image 81 displayed on the monitor 5, the first defect inspection process is automatically selected, and the user selects the second defect inspection processed image displayed on the monitor 5. When 82 is selected, the second defect inspection process is automatically selected. As a result, the search process can be selected by the user's intuitive operation.

The selection of the first defect inspection processed image 81 may be an operation of selecting the first defect inspection processed image 81 itself, an operation of selecting a frame surrounding the first defect inspection processed image 81, and various buttons. It may be a selection operation by such as. The same applies to the selection of the second defect inspection processed image 82. Such an operation can be detected, for example, by detecting the pointer position of the mouse 7 and detecting the click operation. Alternatively, one may be selected from the tool selection drop-down list in the detection level adjustment area 78c of the defect inspection result image display user interface 78 shown in FIG. When the "Next" button in the interface of the detection level adjustment area 78c is pressed after selecting the tool, the tool used for the inspection is confirmed.

(Parameter change in defect inspection processing)
A plurality of defect inspection units 50b shown in FIG. 2 change parameters (parameters for defect inspection) related to image processing applied to inspection images, and repeatedly execute defect inspection processing stored in the inspection setting storage unit 46c. It is configured to get the detection result of. The display control unit 43 is configured to be able to display a plurality of detection results acquired by the defect inspection unit 50b at the same time or by switching.

The defect inspection parameter includes one or more of the presence / absence of filtering, the type of filtering, and the strength of filtering. The defect inspection parameters may include all of the presence / absence of filtering, the type of filtering, and the strength of filtering. The items and values included in the defect inspection parameters may differ between the first defect inspection process and the second defect inspection process.

In the case of scratch inspection tools, the image processing defect inspection parameters applied to the inspection image include parameters that are frequently adjusted when using the scratch inspection tool and pretreatment filters that are compatible with the scratch inspection tool. be able to. A parameter that is frequently adjusted when using a scratch inspection tool is, for example, the size of a "segment". A pretreatment filter that is compatible with scratch inspection tools has "real-time shading correction", and the "extraction size" of this "real-time shading correction" can be used as a parameter for defect inspection.

FIG. 16 shows a list of images obtained by defect inspection processing by changing the parameters related to image processing applied to the inspection image when the tool determined by the inspection processing selection unit 51b is a scratch inspection tool. As an example, the list display interface 93 is shown. In this example, since the tool determined by the inspection process selection unit 51b is the first defect inspection process (scratch inspection tool), the image shown in FIG. 16 is the first defect inspection process image 81, but the inspection process. When the tool determined by the selection unit 51b is the second defect inspection process (blob measurement tool), the second defect inspection process image 82 can be displayed in a list as shown in FIG. Hereinafter, the case where the first defect inspection processed image 81 is displayed will be described, but the same applies to the case where the second defect inspection processed image 82 is displayed.

In the list display interface 93 shown in FIG. 16, the images arranged in the vertical direction are the images in which the "segment size" is changed, and the images arranged in the horizontal direction are the images in which the "extraction size" of the "real-time shading correction" is changed. ing. Since eight parameters are set, eight first defect inspection processing images 81 are displayed. However, the present invention is not limited to this, and two or more parameters may be set. Therefore, the list display interface 93 is displayed. Can display at least two first defect inspection processed images 81. In the list display interface 93, the displayed image can be switched or enlarged and displayed.

The list display interface 93 is provided with a detection level adjustment area 93a. The detection level adjustment region 93a is the same as the detection level adjustment region 78c shown in FIGS. 14 and 15, and the detection level can be adjusted while the first defect inspection processed image 81 is displayed in a list. The adjusted detection level is applied to all first defect inspection processed images 81.

The user can select any one first defect inspection processed image 81 from the plurality of first defect inspection processed images 81 displayed on the list display interface 93. The selection operation of the first defect inspection processing image 81 by the user is accepted by, for example, the operation input unit 44. For example, when the user puts the pointer of the mouse 7 on the image displayed as "2" from the plurality of first defect inspection processed images 81 displayed on the monitor 5 and clicks the image, the selection of the image is completed. do.

The selection of the first defect inspection processed image 81 may be an operation of selecting the first defect inspection processed image 81 itself, an operation of selecting a frame surrounding the first defect inspection processed image 81, and various buttons. It may be a selection operation by such as. Such an operation can be detected, for example, by detecting the pointer position of the mouse 7 and detecting the click operation. Alternatively, "7" may be selected from the image selection drop-down list in the detection level adjustment area 93a of the list display interface 93 shown in FIG. When the "End" button in the interface of the detection level adjustment area 93a is pressed after the image is selected, the parameters used for the inspection are confirmed.

For example, by arranging the "Next" button instead of the "End" button in the detection level adjustment area 93a, the first stage parameter change is performed first, and then the second parameter related to the changed parameter is changed. Detailed setting of parameters may be made possible by making the parameter changes into multiple layers, such as adjusting the parameters of the stage.

In the case of multi-layering, the image selection screen as shown in FIG. 16 appears a plurality of times in the process of image selection. It can also be configured to accept an operation to return to the previous selection screen by pressing the "forward" button in the interface of the detection level adjustment area 93a. By returning to the previous screen while maintaining the parameters selected in the image, it is possible to return to the screen of FIG. 14 and compare the inspection tools with the parameters adjusted to the preferable parameters.

FIG. 17 shows a selection display interface 94 that enlarges and displays the selected first defect inspection processed image 81 for final confirmation after the parameters are determined. The selection display interface 94 is provided with a parameter adjustment area 94a. The parameter adjustment area 94a is provided with an operation button or the like for individually changing the parameters, and the user can change or finely change the parameters as necessary even after selecting the first defect inspection processing image 81. Adjustment is possible.

An image process that enhances the stability of the defect inspection process by combining with the defect inspection process selected by the inspection process selection unit 51b is set in advance, and the image process is applied to the selected inspection image to be subjected to the defect inspection process. It can also be configured to do so automatically.

For example, since the algorithm of the scratch inspection tool is a process of detecting the difference from the periphery, the single algorithm itself detects both white scratches and black scratches on the periphery, and has no light / dark selection function. On the other hand, the "real-time shading correction" installed as a pretreatment filter has a function to sort out the light and dark areas of bright and dark areas with respect to the surroundings, so "real-time shading correction" is combined with a scratch inspection tool. By using it, you can add a light / dark selection function to the scratch inspection tool. The image inspection device 1 appropriately sets the parameters of the preprocessing filter based on the light and dark information of the defect (scratch) that the user inputs in advance, so that there is no knowledge of the algorithm characteristics or the preprocessing filter. However, you can easily create settings that a skilled person would do. The types of pretreatment filters are shown in FIG. 18, but filters other than these filters can also be mounted. The filter type display interface 96 shown in FIG. 18 is for opening a pre-processing filter display unit 96a for displaying the name of the pre-processing filter and a detailed setting screen for the pre-processing filter displayed on the pre-processing filter display unit 96a. A detail button 96b is provided. The pre-processing filter display unit 96a is provided with a drop-down list, and the pre-processing filter can be selected from the drop-down list.

Setting the size of the "segment", which is a parameter of the scratch inspection tool, to the same size as the defect to be detected is the setting that can maximize the detection ability. By appropriately setting the initial value by the image inspection device 1 based on the defect size information input in advance by the user, it is possible to present an appropriate setting based on the characteristics of the algorithm at the initial stage. can.

Since the blob measurement tool is based on simple binarization, binarization is not stable when light and dark shading is applied. On the other hand, the "real-time shading correction" installed as a pre-processing filter has the effect of excluding light and dark shading, so by using the "real-time shading correction" in combination with the blob measurement tool, the blob measurement tool can be used. The binarization can be stabilized.

In addition, the blob measurement tool has the disadvantage of reacting to fine noise, but by combining it with the "blurring process" installed as a preprocessing filter, noise immunity can be greatly improved. The image inspection device 1 appropriately sets the parameter initial values of the preprocessing filter based on the defect light / dark information and the defect size information input in advance by the user, so that there is no knowledge of the preprocessing filter. However, it is possible to easily create settings that a skilled person would make.

In addition, the blob measurement tool has an internal parameter that excludes those with an area below a certain level, but the image inspection device 1 has an initial value based on the defect size information that the user inputs in advance. By setting properly (for example, excluding those with an area less than half the size of the specified defect), it is possible to present the appropriate setting at the initial stage without adjusting the parameters on the user side. can.

It is also good to set the "detection direction" as a parameter to be changed internally for the algorithm of the scratch inspection tool. This is because the parameter in the detection direction is likely to be effective for the work W in which the background pattern is continuous in a uniform direction. Further, in the case of the blob measurement tool, it is preferable to obtain a plurality of detection result images as shown in FIG. 16 by changing the extraction size of the “real-time shading correction” and the intensity of the “blurring process”.

(Explanation of setting mode of image inspection device 1)
Next, the setting mode of the image inspection device 1 will be specifically described. The flowchart shown in FIG. 19 shows the procedure of the setting mode of the search process, and proceeds to step SA1 by a predetermined start operation by the user. In step SA1, the work W is imaged by the image pickup unit 3 to acquire a registered image. do. In step SA2, the registered image 64 acquired in step SA1 is stored in the registered image storage unit 46a and displayed on the monitor 5. After storing and displaying the registered image 64, the process proceeds to step SA3. In step SA3, as shown in FIG. 20, the selection interface 98 is superimposed and displayed on the registered image 64. The selection interface 98 is provided with a "scratch inspection" button 98a and a "search" button 98b so that the user can click and operate with the mouse 7. FIG. 20 shows a state in which the “search” button 98b is clicked.

In step SA3 of FIG. 19, it is determined whether or not the search process is selected. Specifically, it is detected whether or not the "search" button 98b of the selection interface 98 shown in FIG. 20 is clicked and whether or not the "scratch inspection" button 98a is clicked, and the "search" button 98b is clicked. If the result is YES, the process proceeds to step SA4. If the “scratch inspection” button 98a is clicked, the result is determined to be NO, and the steps after step SB4 in the flowchart shown in FIG. 21 to be described later are performed.

In step SA4 of the flowchart shown in FIG. 19, the setting of the angle range of the search process is accepted. Specifically, the common setting item input field 67 shown in FIG. 9 is displayed on the monitor 5, and the user can operate the angle range input unit 67a to set the angle range of the search process.

When it is detected that the setting of the angle range of the search process is completed and the "OK" button is pressed, the process proceeds to step SA5 of the flowchart shown in FIG. 17, and the setting of the search area is accepted. Specifically, as shown in FIG. 6, on the registered image 64 displayed on the monitor 5, the user draws a frame line 64b so as to surround the search area by operating the mouse 7. As a result, the search area is set.

After the search area is set, the process proceeds to step SA6 to accept the setting of the pattern area. Specifically, as shown in FIG. 6, on the registered image 64 displayed on the monitor 5, the user draws a frame line 64a so as to surround the work to be searched by the operation of the mouse 7. As a result, the pattern area is set. The order of steps SA4 to SA6 can be changed.

After passing through steps SA4 to SA6, the process proceeds to step SA7, the first search processing feature amount and the second search processing feature amount are extracted from the image in the pattern region, and stored in the search processing feature amount storage unit 46b. Let me.

After step SA7, the process proceeds to step SA8, and the search processing unit 50a executes the first search process and the second search process. In this example, the first search process is a pattern search and the second search process is an edge-based search.

After executing the search process, in step SA9, the search result image display user interface 68 shown in FIG. 10 is displayed on the monitor 5, and the first search process result image 71 and the second search process result image 72 are displayed on the monitor 5. indicate.

With the first search processing result image 71 and the second search processing result image 72 displayed on the monitor 5, the process proceeds to step SA10, and the detection level of the detection level adjustment region 68c is adjusted. The first search processing result image 71 and the second search processing result image 72 after adjusting the detection level are displayed on the monitor 5 (see FIG. 11). This step SA10 may be performed as needed or may be omitted.

In the subsequent step SA11, the search process selection unit 51a accepts the selection of the search process. After selecting the search process, the process proceeds to step SA12, and the selected search process is stored in the inspection setting storage unit 46c.

After storing the search processing, the process proceeds to step SA13, the search processing parameters are changed, the search processing selected in step SA11 is executed for each parameter, and a plurality of search processing result images are obtained. In the following step SA14, the list display interface 90 shown in FIG. 12 is displayed on the monitor 5, and the search processing result image in step SA13 is displayed on the monitor 5.

In step SA15, the selection of the search processing parameter is accepted. When the user selects one search processing result image from a plurality of search processing result images on the list display interface 90 by operating the mouse 7 or the like, the search processing applied to the selected search processing result image is applied. Read the parameters and store them once.

The search processing result image selected in step SA15 is displayed on the monitor 5 by the selection display interface 91 as shown in FIG. In step SA16, the operation buttons of the parameter adjustment area 91a of the selection display interface 91 are operated to change or fine-tune the parameters. Step SA16 may be performed as needed and may be omitted.

The flowchart shown in FIG. 21 shows the procedure of the setting mode of the defect inspection process, and proceeds to step SB1 by a predetermined start operation by the user. In step SB1, the work W is imaged by the image pickup unit 3 to obtain an inspection image. get. In step SB2, the inspection image acquired in step SB1 is stored in the storage unit 46 and displayed on the monitor 5. After storing and displaying the inspection image, the process proceeds to step SB3. In step SB3, as shown in FIG. 22, the selection interface 98 similar to that in FIG. 20 is superimposed and displayed on the inspection image. FIG. 22 shows a state in which the “scratch inspection” button 98a is clicked.

In step SB3 of FIG. 21, it is determined whether or not the defect inspection (scratch inspection) process is selected. Specifically, it is detected whether or not the "search" button 98b of the selection interface 98 shown in FIG. 22 and whether or not the "scratch inspection" button 98a is clicked, and the "scratch inspection" button 98a is clicked. If this is done, the result is determined to be YES, the process proceeds to step SB4, and if the "search" button 98b is clicked, the result is determined to be NO, and the steps after step SA4 in the flowchart shown in FIG. 19 are performed.

In step SB4 of the flowchart shown in FIG. 21, input of light / dark information is accepted. Specifically, the common setting item input field 66 shown in FIG. 8 is displayed on the monitor 5, and the user operates the drop-down list button 66a to "both black and white", "white scratches only", and "black scratches". You can select one from "only".

When it is detected that the input of the light / dark information is completed and the "OK" button is pressed, the process proceeds to step SB5 of the flowchart shown in FIG. 21, and the setting of the inspection area is accepted. Specifically, as shown in FIG. 7, on the inspection image 65 displayed on the monitor 5, the user draws a frame line 65a so as to surround the inspection area by operating the mouse 7. As a result, the inspection area is set.

After the inspection area is set, the process proceeds to step SB6, and the defect size setting is accepted. Specifically, as shown in FIG. 7, on the inspection image 65 displayed on the monitor 5, the user draws a frame line 65b so as to surround the defective portion 65c by operating the mouse 7. Instead of drawing the frame line 65b, the input with the numerical value of the defect size described above or the input by drawing the defect portion may be performed. This sets the defect size. The order of steps SB4 to SB6 can be changed.

After passing through steps SB4 to SB6, the process proceeds to step SB7, and for each image in the inspection area, the first defect inspection feature amount and the second defect inspection feature amount are extracted and stored in the storage unit 46.

After step SB7, the process proceeds to step SB8, and the defect inspection unit 50b executes the first defect inspection process and the second defect inspection process. In this example, the first defect inspection process is used as a scratch inspection tool, and the second defect inspection process is used as a blob measurement tool.

After executing the defect inspection process, in step SB9, the defect inspection result image display user interface 78 shown in FIG. 14 is displayed on the monitor 5, and the first defect inspection process image 81 and the second defect inspection process image 82 are displayed on the monitor. Display in 5.

With the first defect inspection processed image 81 and the second defect inspection processed image 82 displayed on the monitor 5, the process proceeds to step SB10, and the detection level of the detection level adjustment region 78c is adjusted. The first defect inspection processed image 81 and the second defect inspection processed image 82 after adjusting the detection level are displayed on the monitor 5 (see FIG. 15). This step SB10 may be performed as needed or may be omitted.

In the subsequent step SB11, the inspection process selection unit 51b accepts the selection of the defect inspection process. After selecting the defect inspection process, the process proceeds to step SB12, and the selected defect inspection process is stored in the inspection setting storage unit 46c.

After storing the defect inspection process, the process proceeds to step SB13, the parameter for defect inspection is changed, the defect inspection process selected in step SB11 is executed for each parameter, and a plurality of defect inspection processing images are obtained. In the subsequent step SB14, the list display interface 93 shown in FIG. 16 is displayed on the monitor 5, and the defect inspection processed image of the step SB13 is displayed on the monitor 5.

In step SB15, the selection of the defect inspection parameter is accepted. When the user selects one defect inspection processed image from a plurality of defect inspection processed images on the list display interface 93 by operating the mouse 7 or the like, the defect inspection applied to the selected defect inspection processed image is applied. Read the parameters and store them once.

The defect inspection processed image selected in step SB15 is displayed on the monitor 5 by the selection display interface 94 as shown in FIG. In step SB16, the operation buttons in the parameter adjustment area 94a of the selection display interface 94 are operated to change or fine-tune the parameters. Step SB16 may be performed as needed and may be omitted.

(Action and effect of the embodiment)
As described above, according to the image inspection apparatus 1 according to this embodiment, the result of the first search processing using the first search processing feature amount extracted from the image in the pattern region used for the search processing. The obtained first search processing result image and the second search processing result image obtained as a result of the second search processing using the second search processing feature amount are displayed on the monitor 5 to easily check the detection state. Since comparisons can be made, even a user who does not know the characteristics of a plurality of search algorithms and compatibility with the work W can easily determine an appropriate search algorithm. In addition, it is possible to set appropriate parameters based on the algorithm characteristics and effective preprocessing filter combinations based on the specified information, so even users who do not know the algorithm characteristics are appropriate. Parameters can be easily determined.

Further, a first defect inspection processed image showing a defect area obtained as a result of the first defect inspection process using the first search processing feature amount in the inspection area used for the defect inspection process, and a second search. Since it is possible to easily compare the detection states by displaying the second defect inspection processing image showing the defect area obtained as a result of the second defect inspection processing using the feature quantity for processing, a plurality of defect inspection algorithms. Even a user who does not know the characteristics of the image or compatibility with the inspection object can easily determine an appropriate defect inspection algorithm. In addition, it is possible to set appropriate parameters based on the algorithm characteristics and effective preprocessing filter combinations based on the specified information, so even users who do not know the algorithm characteristics are appropriate. Parameters can be easily determined.

The above embodiments are merely exemplary in all respects and should not be construed in a limited way. Further, all modifications and modifications belonging to the equivalent scope of the claims are within the scope of the present invention.

As described above, the image inspection apparatus according to the present invention can be used in the field of inspecting the appearance of a work such as various products.

1 Image inspection device 2 Lighting unit 3 Imaging unit 4 Control unit 5 Monitor 6 Keyboard 7 Mouse 43 Display control unit 46a Registered image storage unit 46b Search processing feature quantity storage unit 46c Inspection setting storage unit 47a Pattern area setting unit 47b Inspection area setting Unit 47c Defect size setting unit 48 Defect inspection feature quantity extraction unit 49 Item input unit 50a Search processing unit 50b Defect inspection unit 51a Search processing selection unit 51b Inspection processing selection unit

Claims (12)

In an image inspection device that inspects the appearance of an object to be inspected,
A registered image storage unit that stores a registered image obtained by imaging the inspection object, and a registered image storage unit.
A pattern area setting unit for setting a pattern area used for search processing on the registered image stored in the registered image storage unit, and a pattern area setting unit.
A feature amount storage unit that stores a shading image extracted from an image in the pattern area and edge information , and a feature amount storage unit.
Normalization, which is a first search process for detecting a first image region having a first correlation value equal to or higher than a predetermined value with the grayscale image for an input image newly obtained by imaging the inspection object. A search processing unit that executes a normalization correlation search , an edge-based search that is a second search process for detecting a second image region having a second correlation value equal to or higher than a predetermined value with the edge information , and a search processing unit.
The position of the first image region obtained as a result of the normalized correlation search , the first image displaying the first correlation value of the first image region, and the said image obtained as a result of the edge-based search . A display control unit capable of displaying the position of the second image area and the second image displaying the second correlation value of the second image area at the same time or by switching.
A search processing selection unit that accepts selection of either the normalized correlation search or the edge-based search after the first image or the second image is displayed.
An image inspection device including an inspection setting storage unit that stores the search process selected by the search processing selection unit as inspection settings. In the image inspection apparatus according to claim 1,
The display control unit has an image having a correlation value of the highest correlated image region having the highest correlation value and a correlation value less than the correlation value of the highest correlated image region among a plurality of image regions having a correlation value of a predetermined value or more. An image inspection device configured to display an area in a different display form. In the image inspection apparatus according to claim 1 or 2.
The display control unit displays a detection frame surrounding an image area having a correlation value equal to or higher than the predetermined value, and the larger the size of the correlation value, the larger the detection frame. Device. In the image inspection apparatus according to claim 2,
An image inspection device configured to be able to adjust the predetermined value. In the image inspection apparatus according to any one of claims 1 to 4.
The search processing selection unit detects whether or not the first image or the second image displayed by the display control unit has been selected, and if it detects that the first image has been selected, the search process selection unit may detect that the first image has been selected. An image inspection apparatus configured to select a normalized correlation search and select the edge-based search when it detects that the second image has been selected. In the image inspection apparatus according to any one of claims 1 to 5.
The search processing unit acquires the processing time of the normalized correlation search and the processing time of the edge-based search .
The display control unit is an image inspection device configured to display the processing time of the normalized correlation search acquired by the search processing unit and the processing time of the edge-based search . In the image inspection apparatus according to any one of claims 1 to 6.
The search processing unit continuously acquires a plurality of the input images and executes the normalized correlation search and the edge-based search on each of the input images.
The display control unit is an image inspection device configured to update the first image and the second image. In the image inspection apparatus according to any one of claims 1 to 7.
It is provided with an item input unit for inputting common setting items of the normalized correlation search and the edge-based search .
The search processing unit is an image inspection device configured to apply the common setting items input by the item input unit to the normalized correlation search and the edge-based search . In the image inspection apparatus according to claim 8,
The common setting item is an image inspection device that is an angle range for performing search processing. In the image inspection apparatus according to any one of claims 1 to 9.
The search processing unit is configured to change parameters related to image processing applied to the input image, repeatedly execute the search processing stored in the inspection setting storage unit, and acquire a plurality of detection results. ,
The display control unit is an image inspection device configured to be capable of displaying a plurality of detection results acquired by the search processing unit at the same time or by switching. In the image inspection apparatus according to claim 10,
The image inspection apparatus includes any one of the presence / absence of filtering, the type of filtering, and the intensity of filtering in the parameters. In the image inspection apparatus according to claim 10 or 11.
An image inspection device whose parameters include the degree of image reduction.

Images