JP2003139522A - Image inspection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a convenient inspection system by reducing the burden on workman, eliminating individual difference or the like of judgement, and providing a specification changing function, a resolution automatic calculating function and an inspection object image segmenting function of an image inspection device. SOLUTION: A positioning machine is provided for selecting an inspection portion by X and Y stages and a projection angle setter while looking at a screen of a CCD camera. Such a means is provided that is operating in agreement with image processing software designating inspection contents at a position of the inspection portion and a script file format description describing operation procedures such as lighting control. A means is provided for automatically calculating resolution of the camera. A segmented image is inspected until uniformity of measurement conditions is secured.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image inspection apparatus for automatically detecting an abnormality in image data taken by a camera, which has a function of easily creating and changing an inspection procedure and an automatic correction function of resolution, which is easy to use. Suggest a system.

[0002]

2. Description of the Related Art Conventionally, the visual inspection of products has been mainly conducted by human eyes. The appearance inspection is to check whether the product is scratched, dust is not attached, dirt is not present, color is not uneven, different parts are not attached, or the like.
Sometimes the inspection is done by devising lighting so that it is easy to see, or by attaching a magnifying lens. Mechanization of this has also been considered and is partly well implemented. Specifically, it consists of a camera, a driving device such as XY, and computer software.

[0003]

Problems to be Solved by the Invention The problems to be solved by the present invention are as follows. (1) In many cases, the work to be done by a person is small,
There were problems such as the burden on workers due to fatigue, oversight, and individual differences in judgment. (2) The mechanization is only partially spread, and it is impossible to raise the performance to a truly effective level with a combination of general-purpose devices. It has been pointed out that it is expensive to build a dedicated system, and it is difficult to change specifications and reuse it in other products.

[0004]

[Means for Solving the Problems] In order to solve the above problems, the following means are proposed. (1) First, move the X and Y stages to the position P1 and inspect at the angle θ1. Next, the X and Y stages are moved to the position P2 and inspected at the angle θ2. Next, a means is provided that allows the user to select a means for determining the position of the drive system in the form of ... (2) The image processing software for designating the inspection contents at the position of (1) above and means for operating in accordance with the description in the script file format which describes the operation procedure such as lighting control are provided. (3) A means for automatically calculating the resolution of the camera is provided. (4) The subdivided image is inspected until the uniformity of the measurement conditions is ensured.

[0005]

BEST MODE FOR CARRYING OUT THE INVENTION In the present invention, a mechanism for easily changing the specifications of an image inspection apparatus is provided, a function for automatically calculating resolution is provided, and an image to be inspected is subdivided so that the measurement conditions are uniform. By providing such a means, it was possible to realize a system with high productivity and ease of use.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a block diagram of an embodiment of the present invention (an example of an image inspection apparatus using a CCD camera). 1 is the inspection object 4
It is a camera for photographing, and can be moved in the direction of arrow 3 by the linear drive system 2. This mainly works to secure the focal length of the camera. On the table on which the inspection object is placed, the shooting angle of the camera can be changed by the θ-axis motor 5. Reference numeral 6 denotes a linear drive system for moving in the X-axis direction 7, which works to select a photographing position of the inspection object. Reference numeral 8 denotes a light source that illuminates the back surface of the inspection object. Although not shown, it goes without saying that an illumination means from the front surface of the inspection object can be provided.

A method for finding dust in an image by image processing will be described with reference to FIG. Scratches and dust have different colors and densities with respect to the surrounding colors. A normal image fits within a predetermined normal distribution, but scratches and dust deviate from these to find defects. A specific example will be described. Consider a case where an image to be inspected is dark in gray and a dust image is present as shown in FIG. A histogram of the density data in this case is shown in FIG.
It becomes like -2. An average value of 120.75 and a standard deviation of 3.91 were calculated. For this image, there are 3000 pixels since there are 60 * 50 pixels. It is expected that almost all normal images will exist in the 3.5σ range (120.75-3.5 * 3.91 to 120.75 + 3.5 * 3.91), so if you create a black image for pixels outside this range and a white image for others, it will be as shown in Figure 2- It will be 3. With this, only the dust position in the center is painted black, and as a result, dust can be extracted. The size of dust can be known by counting the number of pixels of the black dots.

FIG. 3 is a screen for inputting the operating procedure of the drive system. Although not shown in FIG. 1, the output image of the camera is displayed on the monitor in real time. The user presses the X-axis, Z-axis, and θ-axis motor movement keys on the computer GUI at the position to be inspected while looking at the camera screen, and if OK, presses the enter key to register. By repeating this, the operating procedure of the drive system can be registered.

FIG. 4 illustrates the function of calculating the image resolution. In this embodiment, the test image is an image with one black dot. It is assumed that there is a black dot at the position shown in Fig. 4-1 at first. For example, assume that the X-axis drive system is moved by L inches and the subsequent image is shown in FIG. Calculate the center of gravity of the black dots in Figures 4-1 and 4-2. Here, the method of calculating the center of gravity (X0, Y0) is to calculate the coordinates of each of the n pixels forming the black dot as (Xi, Yi).
Can be calculated as X0 = {ΣXi} / n and Y0 = {ΣYi} / n. If the coordinates in FIG. 4-1 are (x1, y1) and the coordinates in FIG. 4-2 are (x2, y2), the spatial distance on the pixel is √ {(x1-x2) ^ 2 + (y1-y
2) ^ 2} = L inches, and the image resolution DPI = √ {(x1
-It can be calculated by x2) ^ 2 + (y1-y2) ^ 2} / L. The resolution setting is also the same as the above "Drive system measurement condition trace function", specify it on the screen and move it to the (x1, y1) position, then "Set", and move it to the (x2, y2) position. Pressing "Set" will automatically calculate and set the resolution.

Next, a method for describing the entire inspection procedure will be described. For that purpose, it is necessary to divide the contents of image processing into parts. To make it into a component means to divide and prepare software for each function. That way, every time the inspection object changes, the program will be rewritten,
Productivity decline can be avoided. The image inspection basically consists of movement of an object to be inspected (control of drive system), image capturing, image processing and determination, and output of results. It is lighting control if added. If you create a platform that can combine these into parts, the versatility will increase. The general processing procedure can be described in a script file format to enhance versatility. An example is shown below.

[0011] Script format description example -------------------------------------------------- ---------------------- -------------------------- START Light 1 on, Light 2 off: Turn on 1 of the 2 light sources and turn off 2 of them. Move1: Moves to 1 specified by the measurement position trace function. Capture size1, param1, param2: Capture image with specified size and parameter. Img_process1: Specified image processing, judgment, and output of results. Move2: Move to 2 specified by the measurement position trace function. Capture size1, param1, param2: Capture image with specified size and parameter. Img_process1: Specified image processing, judgment, and output of results. Light 1 off, Light 2 on: Turn off 1 of the 2 light sources and turn on 2. Move3: Move to 3 specified by the measurement position trace function. Capture size2, param1, param2: Capture image with specified size and parameter. Img_process2: Specified image processing, judgment, and output of results. END -------------------------------------------------- ---------------------- --------------------------

Attempts to process the entire image with the same criteria may not work. This is because the measurement conditions are different on the upper right and lower left of the screen, and thus the same criterion may cause inconsistency. For example, the measurement conditions are
The amount of light and the focus. In order to solve this problem, it is effective to divide the processing into small spaces with almost the same conditions and perform processing. I read this small space as a window. FIG. 5 shows a case where one screen is divided into four and image processing is performed.

It is also important to apply a plurality of processes in multiple stages. Accurate judgment cannot be made with one algorithm and one criterion. Therefore, it is important to find out the problems with multiple algorithms (criteria) so as not to overlook the problems. The idea is to find out 30% by inspection A, 25% by inspection B, 25% by inspection C, and 20% by weight inspection D. The concept is shown in FIG.

[0014]

[Brief description of drawings]

FIG. 1 is a configuration example of the present invention.

FIG. 2 Example of image processing

[Figure 3] Screen example for tracing the measurement position

FIG. 4 is an explanatory diagram of resolution measurement.

FIG. 5 is an explanatory diagram of image division processing.

FIG. 6 is an explanatory diagram of multistage image processing.

[0015]

Continued front page    F term (reference) 2F065 AA17 AA49 AA58 DD03 DD04                       FF04 GG09 HH16 JJ03 JJ26                       KK03 MM03 MM04 MM07 PP12                       QQ03 QQ42 QQ43 QQ51 SS13                       UU05                 2G051 AB01 AB02 AC02 AC15 CA04                       DA07 EC02 ED09 ED23                 3F079 CA31 CA41 CB29 CB37 EA16                 5B057 AA01 BA02 BA24 DA03 DA20                       DB02 DB09 DC05 DC06 DC30                       DC32 DC36

Claims (4)

[Claims] 1. A camera for reading an image of an object to be inspected, a drive system for changing a measuring position, and an image for determining whether the object to be inspected is good or bad by taking an image from the camera and performing a predetermined image processing. The inspection apparatus is characterized by having a function of designating the position of the drive system on the screen of the control computer and a function of tracing the measurement conditions of the drive system by means of recording the confirmed position. 2. A script file in which the entire image processing software is decomposed for each function, and at least the driving condition of the above 1, the software for each function, and the operation procedure of three items of illumination control for the object to be inspected are described. An image inspection device characterized in that it operates according to the description of the format. 3. A camera for reading an image of an object to be inspected, a drive system for changing a measurement position, an image from the camera and performing predetermined image processing to determine whether the object to be inspected is good or bad. In the inspection apparatus, as a means for calculating the resolution of an image captured from the camera, a test image is placed and the image is captured (image A), and a test image is placed after a known predetermined amount (L) movement and the image is captured (image B), an image inspection apparatus characterized in that the resolution of the captured image is calculated from the images A, B and L. 4. A camera for reading an image of an object to be inspected, a drive system for changing a measuring position, and an image for determining whether the object to be inspected is good or bad by taking an image from the camera and performing a predetermined image processing. The inspection apparatus is characterized in that the image of the object to be inspected is divided into a plurality of images and the subdivided images are inspected until the uniformity of the measurement conditions is ensured.