Fault Location Device and Fault Location Method

Abstract

To provide a fault position identification device with which, in a device for identifying the position of a fault having occurred to the design surface of a vehicle body which is the object to be detected, it is possible to calculate the movement amount of a transported vehicle body from imaged data obtained by imaging means and identifying the position of the fault on the basis of the movement amount, allowing for highly accurate position identification, and a fault position identification method using this fault position identification device.SOLUTION: A fault position identification device comprises design surface imaging means for imaging the design surface of a vehicle body, lower bottom surface imaging means for imaging a lower bottom surface which is a non-design surface of the vehicle body, and information processing means for processing information obtained from the design surface imaging means and the lower bottom surface imaging means.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a defect position specifying device and a defect position specifying method for specifying the position of defects such as scratches and painting unevenness occurring on a design surface of an object to be inspected such as a vehicle body moved by a conveying device.

Background Art

[0002] Conventionally, among vehicles manufactured in a factory, after the painting of the design surface of the vehicle body is completed, while being moved by a conveying device such as a chain conveyor, an inspection is performed to check whether there are any defects such as scratches and painting unevenness on the design surface. Specifically, the design surface of the vehicle body is photographed using imaging means disposed in the vicinity of the conveying device, and the presence or absence of defects is confirmed based on the photographed image. If a defect occurs, it was specified by calculating the number of rotations of the motor constituting the conveying device from the head portion of the vehicle body to the position where the defect occurred using an encoder. However, in the conventional defect position specifying method, it is necessary to temporarily stop the conveying device at the imaging start point by the imaging means in order to determine the measurement start position, and also to temporarily stop the conveying device to measure the number of rotations when a defect is detected. Therefore, there is a drawback that the cycle time related to the manufacturing process is extended by stopping the conveying device each time. In addition, since the amount of movement is calculated based on the number of rotations of the motor of the conveying device to specify the position of the defect, it is not possible to specify the position of the defect in detail, and there is also a problem that it takes time for an operator to visually confirm the defect.

[0003] Patent Document 1 describes a surface defect detection device and a detection method for detecting defects occurring on the surface based on a plurality of images obtained when continuously imaging while irradiating a measurement portion of a vehicle moved by a conveying device with pattern light using imaging means. According to the technique described in this Patent Document 1, a plurality of sensors such as a position sensor and a speed sensor are provided, and detection of the entry of the vehicle into the imaging area by the imaging means and measurement of the speed at which the vehicle is conveyed are performed.

Prior Art Documents

Patent Document

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] According to the surface defect detection device and detection method described in Patent Document 1, while irradiating the painted surface of the vehicle with the illumination light of the light and dark pattern, the imaging means is used to continuously image the design surface, and based on the obtained plurality of images, accurate defect detection can be performed by detecting defects. Regarding the identification of the position of the detected defect, it is described that it is made to correspond to the position information of the vehicle body detected by a position sensor or a speed sensor. And regarding the vehicle position detection method, various methods are described, such as a method combining a sensor for detecting the passage of a reference point such as an imaging start point and a distance sensor, and a method combining a sensor for detecting the passage of a reference point, a speed sensor for measuring the traveling speed of the vehicle, and the interval between the consecutive imaging times of the front and rear adjacent images continuously acquired by the imaging means. However, the position information of the vehicle body linked to the identification of the defect position is measured by using various sensors, which requires a plurality of sensors for vehicle position detection, resulting in complicated processing, and there is a risk of high introduction and maintenance costs.

[0006] Also, Patent Document 1 describes that in the continuously captured images, the amount of positional deviation between the front and rear images is calculated and synthesized from the speed of the vehicle. By synthesizing this composite image from the front of the vehicle, it is also conceivable to confirm the distance from the front of the vehicle to the defect on the screen. However, since there is no method for converting the distance on the screen into the actual distance, it can only be confirmed on the screen. In addition, since the method for calculating the described misalignment amount is calculated from the vehicle speed, sensors such as a speed sensor that directly measures the vehicle speed and a position sensor for calculating the speed from the position are required. And, since the misalignment amount is calculated based on the traveling speed of the vehicle, it is difficult to cope with cases where the conveying device stops or the speed varies, and there is a risk that the accuracy of image synthesis will decrease.

[0007] The present invention has been made in view of the above problems, and by imaging the lower bottom surface of the vehicle body, which is the inspection object, it enables image synthesis only with the features reflected in each of the front and rear images, and aims to provide a defect position specifying device and a defect position specifying method capable of calculating the amount of movement (number of pixels) of the vehicle body on the screen.

Means for Solving the Problems

[0008] In order to solve the above conventional problems, in a defect position specifying device, there are provided a design surface imaging means for imaging the design surface of the vehicle body, a lower bottom surface imaging means for imaging the lower bottom surface which is a non-design surface of the vehicle body, and an information processing means for processing the information obtained from the design surface imaging means and the lower bottom surface imaging means. The information processing means includes a defect detection unit for detecting a defect occurring on the design surface of the vehicle body based on the design surface imaging data acquired by the design surface imaging means, decomposes the lower bottom surface imaging data acquired by the lower bottom surface imaging means into a plurality of image frames, and generates a reference image for specifying the position of the defect by connecting the front and rear of the image frames based on a landmark which is an arbitrary image portion included in the image frame. It is characterized by having a defect position specifying unit for specifying the position of the defect detected from the design surface imaging data by the defect detection unit based on the reference image.

[0009] In addition, in the method for identifying the position of a defect, design surface imaging data and bottom surface imaging data are obtained by imaging the design surface and the bottom surface of the vehicle body. The design surface imaging data detects defects based on the acquired information. The bottom surface imaging data is decomposed into a plurality of image frames, and a reference image is generated by connecting the front and rear of the image frames based on a landmark, which is an arbitrary image portion included in the image frame. The position of the defect is calculated based on the design surface imaging data and the reference image, and the distances on the design surface imaging data and the bottom surface imaging data are converted into actual distance units to identify the position of the defect, which is also a feature of this method.

Advantages of the Invention

[0010] According to the present invention, by imaging the bottom surface of the vehicle body, which is the object to be inspected, it is possible to generate a reference image by overlapping and synthesizing the same imaging range using, as landmarks, bolt holes, welding marks, etc. reflected in the front and rear image frames, and based on this reference image, the position of the defect generated on the design surface can be identified.

[0011] In addition, for generating the reference image used for identifying the position of the defect, it is possible to use only the bottom surface imaging data obtained from the bottom surface imaging means, so the position of the defect can be identified without being affected by the conveyance speed of the conveyance device. Furthermore, even if the conveyance device is temporarily stopped during the detection of the defect, the generation of the reference image can be continued from the stop position as it is.

[0012] Also, since the information processing means holds, as master information in advance, the actual distance (mm) per pixel, the position of the defect on the screen can be calculated in terms of the number of pixels, and this value can be converted into a distance unit (mm) based on the master information, enabling the identification of the actual distance from the head of the vehicle body or an arbitrary position on the vehicle body to the defect.

[0013] In addition, the actual height of the defect can be identified by calculating the height of the defect in terms of the number of pixels from the information of the design surface imaging data and converting this value into an actual distance unit (mm) based on the master information.

[0014] Also, by calculating the position on the defect screen in terms of the number of pixels and converting it into the actual distance unit (mm) based on the master information, the distance and height of the defect position from the start can be known, and the defect position can be specified with high accuracy.

[0015] In addition, the defect position specifying apparatus and the defect position specifying method of the present invention can be realized by adding a bottom surface imaging means and an information processing means for processing the information obtained from the bottom surface imaging means to an existing production line, and can be introduced at low cost.

[0016] Also, when synthesizing the bottom surface imaging data to generate a reference image, based on the overlapping information between the front and rear image frames, the actual movement amount per frame of the vehicle body conveyed by the conveying device can be calculated, and the position of the vehicle body on the production line can be specified based on that information.

Brief Description of the Drawings

[0017]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Mode for Carrying Out the Invention

[0018] The gist of this invention is that in a defect position specifying device, there are provided a design surface imaging means for imaging the design surface of a vehicle body, a bottom surface imaging means for imaging the bottom surface which is a non-design surface of the vehicle body, and an information processing means for processing information obtained from the design surface imaging means and the bottom surface imaging means. The information processing means includes a defect detection unit for detecting a defect occurring on the design surface of the vehicle body based on design surface imaging data acquired by the design surface imaging means, a reference image generation unit for decomposing the bottom surface imaging data acquired by the bottom surface imaging means into a plurality of image frames and generating a reference image for specifying the position of a defect by connecting the front and back of the image frames based on a landmark which is an arbitrary image portion included in the image frame, and a defect position specifying unit for specifying the position of the defect detected from the design surface imaging data by the defect detection unit based on the reference image.

[0019] Further, in the method of specifying the position of a defect, design surface imaging data and bottom surface imaging data are acquired by imaging the design surface and the bottom surface of the vehicle body. The design surface imaging data is used to detect a defect based on the acquired information. The bottom surface imaging data is decomposed into a plurality of image frames, and a reference image is generated by connecting the front and back of the image frames based on a landmark which is an arbitrary image portion included in the image frame. The position of the defect is calculated based on the design surface imaging data and the reference image, and the distance on the design surface imaging data and the bottom surface imaging data is converted into actual distance units to specify the position of the defect, which is also a feature.

[0020] Hereinafter, an embodiment of a defect position specifying apparatus and a defect position specifying method according to the present invention will be described with reference to the accompanying drawings. The following embodiments are an example of embodying the present invention and do not limit the technical scope of the present invention. In FIG. 1(b), arrow X indicates the conveyance direction of the conveyance device C.

[0021] In the manufacture of a vehicle, the vehicle body B is conveyed by the conveyance device C after the painting of the design surface is completed, and an inspection is performed to check whether there are any defects such as scratches or painting unevenness on the design surface. The defects detected during the inspection are visually confirmed by an operator at the defect location, and repair work is performed according to the defects.

[0022] In the defect position specifying apparatus and the defect position specifying method according to the present embodiment, when performing the above-described defect detection and defect position specification, it is detected whether there are any defects such as scratches or painting unevenness on the painted vehicle body B being conveyed by the conveyance device C, and the position of the detected defective vehicle body B is specified. That is, according to the defect position specifying apparatus and the defect position specifying method according to the present embodiment, it is possible to perform defect detection and defect position specification for the vehicle body B without stopping the operation of the conveyance device C.

[0023] Further, according to the defect position specifying apparatus and the defect position specifying method according to the present embodiment, even when the conveyance device temporarily stops during the defect detection, it is possible to continue the defect detection and the defect position specification from the stop position.

[0024] Further, according to the defect position specifying apparatus and the defect position specifying method according to the present embodiment, it is possible to calculate the amount of movement of the conveyed vehicle body per frame in the generation of the reference image required for defect position specification, and to specify the position of the vehicle body on the production line based on that information.

[0025] The conveying device C conveys the vehicle body B, which is the object to be inspected, in a certain direction (the direction of arrow X in FIG. 1). In the present embodiment, as shown in FIG. 1(a), a conveying lane narrower than the vehicle body B is used. This is because the bottom surface imaging means 200 described later can image the bottom surface near the side door of the vehicle body B from below. That is, the conveying device C may have a configuration that conveys in a certain direction at a certain speed and allows the bottom surface imaging means 200 to image the bottom surface of the vehicle body B from below.

[0026] The vehicle body B, which is the object to be inspected, is conveyed in a certain direction by the conveying device C. The vehicle body B is conveyed in a state where the design surface is painted, and defects such as scratches and uneven painting occurring on this design surface are the objects to be detected. Also, the bottom surface of the vehicle body B is a non-design surface, and bolt holes, welding marks, etc. for assembling and manufacturing the vehicle body B can be visually recognized.

[0027] [Regarding the device for detecting defects and specifying positions] As shown in FIG. 1, the method for detecting defects and specifying the positions of defects of the vehicle body according to the present embodiment is performed using a defect position specifying device M including a design surface imaging means 100 that images the design surface of the vehicle body B conveyed by the conveying device C, a bottom surface imaging means 200 that images the bottom surface, which is the non-design surface of the vehicle body B, and an information processing means 300 that processes information obtained from the design surface imaging means 100 and the bottom surface imaging means 200.

[0028] The design surface imaging means 100 is composed of an imaging device such as a video camera equipped with an imaging element such as a CCD (Charge Coupled Device). As shown in FIG. 1, the design surface imaging means 100 images the design surface of the vehicle body B, which is the object to be inspected, and transmits the imaged information to the information processing means 300. This information is used to detect whether there are any defects on the design surface. Only one design surface imaging means 100 is shown in FIG. 1 for explaining the present embodiment, but a plurality of such design surface imaging means 100 may be arranged corresponding to the design surfaces to be inspected.

[0029] The bottom surface imaging means 200 is composed of an imaging device such as a video camera, similar to the design surface imaging means 100. The bottom surface imaging means 200 photographs the bottom surface, which is a non-design surface of the vehicle body B, from below the vehicle body B and transmits the photographed information to the information processing means 300. This information is used to generate a reference image 400 for identifying the position of a defect described later.

[0030] As shown in FIG. 1, the design surface imaging means 100 and the bottom surface imaging means 200 are arranged such that the distances from the leading end of the vehicle body B conveyed in the direction of arrow X are the same.

[0031] Based on the information received from the design surface imaging means 100, the information processing means 300 detects defects on the design surface of the vehicle body B and identifies the positions of the detected defects by synchronizing the images photographed by the design surface imaging means 100 and the bottom surface imaging means 200. The information processing means 300 includes, for example, a CPU (Central Processing Unit) as an arithmetic processing device that executes various arithmetic processes and controls, storage devices such as a RAM (Random Access Memory) and a ROM (Read Only Memory), input / output devices such as an input / output interface for data input / output, and peripheral circuits such as a clock circuit, which are connected by a bus or the like. The CPU of the information processing means 300 performs arithmetic processing according to various programs stored in the ROM or the like. In addition, master information indicating how many millimeters each pixel is actually in the images obtained in advance by the design surface imaging means 100 and the bottom surface imaging means 200 is recorded in the information processing means 300.

[0032] As shown in FIG. 2, the information processing means 300 includes a defect detection unit 310, a reference image generation unit 320, and a defect position identification unit 330.

[0033] The defect detection unit 310 detects defects that occur on the design surface of the vehicle body B. Specifically, it compares the design surface imaging data 110 transmitted from the design surface imaging means 100 with the imaging data that records a vehicle body of the same vehicle type without defects, which has been registered in the defect detection unit 310 in advance, to detect defects. Examples of this comparison method include irradiating light on the design surface and detecting defects from the comparison of the reflections. This method has the effect of improving the detection accuracy of paint unevenness, which is particularly difficult to determine in the comparison of data captured in the normal way. However, the specific detection method of the defect detection unit 310 is only an example, and there is no particular limitation regarding how to analyze the design surface imaging data 110 transmitted from the design surface imaging means 100 to detect defects.

[0034] The reference image generation unit 320 generates an image that serves as a reference for specifying the position of a defect that has occurred on the design surface by the defect position specifying unit 330 described later. Specifically, it decomposes the bottom surface imaging data 210 transmitted from the bottom surface imaging means 200 frame by frame, and generates a reference image 400 in which the same locations reflected in the front and rear frames are overlapped based on the landmarks reflected in each frame. By using the bottom surface imaging data 210 that has captured the bottom surface of the vehicle body B to generate this reference image 400, characteristic locations such as bolt holes and welding marks that serve as landmarks are necessarily reflected in each frame image. Therefore, it is possible to determine the same location based on the landmarks reflected in the front and rear frames and overlap the images from the head to the tail of the vehicle body B.

[0035] The defect position specifying unit 330 specifies the position of a defect detected from the design surface imaging data 110 based on the reference image 400 generated by the reference image generation unit 320. Specifically, it calculates the number of pixels from the head of the vehicle body B to the position of the detected defect. Then, the position of the defect can be specified by converting it to the actual distance unit (mm) based on the master information on how many millimeters each pixel is in the defect position specifying unit 330, which has been registered in advance. However, as for more specific methods for identifying the location of the defect in the defect location identification unit 330, multiple methods can be considered. Therefore, the more specific method for identifying the location of the defect will be described in "Regarding the Method for Detecting and Identifying Defects" described later.

[0036] Moreover, the information processing means 300 does not necessarily consist of a single information processing device, and may be an assembly of a plurality of information processing devices separately equipped with the defect detection unit 310, the reference image generation unit 320, and the defect location identification unit 330.

[0037] [Regarding the Method for Identifying the Location of Defects] Next, an example of a series of processes for detecting and identifying defects using the defect location identification device M according to the present embodiment will be described based on the drawings. FIG. 3, FIG. 6, and FIG. 7 are all flowcharts showing the procedure of a method for calculating the distance from the leading end of the vehicle body B to the defective portion 121 using the defect location identification device M.

[0038] First, the distance (the distance in the horizontal direction) from the leading end of the vehicle body B to the defective portion 121 is specified. As shown in FIG. 3, in the method for detecting and identifying defects, when the vehicle body B conveyed by the conveying device C enters the imaging area, the design surface imaging means 100 and the bottom surface imaging means 200 start imaging simultaneously. At this time, as described above, the design surface imaging means 100 and the bottom surface imaging means 200 are arranged so that the distances from the leading end of the vehicle body B are the same (see FIG. 1).

[0039] As shown in FIG. 3, the design surface imaging means 100 captures the painted design surface of the vehicle body B, which is the inspection object, and transmits the design surface imaging data 110 to the defect detection unit 310 of the information processing means 300.

[0040] The defect detection unit 310 holds, as comparison imaging data, a body image without defects that is the same as the vehicle model of the inspection target recorded in advance. The defect detection unit 310 that has received the design surface imaging data 110 compares the image of the comparison imaging data with the design surface imaging data 110 to detect a defect by checking whether there is a defect in the design surface of the vehicle body B.

[0041] When a defect is detected, the image frame that captured the defect is recognized as a defect image 120 and the defect location 121. Calculate from which frame the recognized defect image 120 is the frame that started imaging the front of the vehicle body B. Information about which frame it is is transmitted to the defect position specifying unit 330 of the information processing means 300.

[0042] Any method may be used for this method of detecting a defect in the design surface as long as it can check whether there are scratches or uneven painting on the design surface of the vehicle body B, which is the inspection target. For example, a method of detecting by irradiating light and detecting an abnormality in the reflection may be used.

[0043] As shown in FIG. 3, the bottom surface imaging means 200 photographs the bottom surface, which is the non-design surface of the vehicle body B, which is the inspection target, and transmits the bottom surface imaging data 210 to the reference image generation unit 320 of the information processing means 300.

[0044] Incidentally, the bottom surface is the part on the back side with respect to the design surface of the vehicle body B, that is, the vehicle body surface. The bottom surface of this vehicle body B is in a state where bolt holes for assembling parts of the interior surface of the automobile and press marks and welding marks made during the manufacture of the vehicle body B can be visually recognized. Therefore, the bolt holes and processing marks existing on the bottom surface will function as marks 230 for image splicing and synthesis performed by the reference image generation unit 320.

[0045] The reference image generation unit 320 of the information processing means 300 decomposes the bottom surface imaging data 210 received from the bottom surface imaging means 200 frame by frame, and combines them to generate a reference image 400 that serves as a reference in identifying the position of defects.

[0046] As shown in FIG. 4, the specific method for generating the reference image 400 is to use the bolt holes and machining marks reflected in the image frames 220a to 220z obtained by decomposing the bottom surface imaging data 210 frame by frame as marks 230, and overlap and connect the same imaging ranges of the front and rear frames. That is, the mark 230 is an arbitrary image part included in the image frames 220a to 220z and is set when generating the reference image 400. As the mark 230, a special part that is clearly distinguishable from the general part, which is a monotonous image part, is used in each of the image frames 220a to 220z. The mark 230 is, for example, a characteristic image part such as bolt holes and machining marks reflected in each of the image frames 220a to 220z. Specifically, when the mark 230 is a bolt hole, the front and rear image frames 220 are overlapped so that the positions of the bolt holes reflected in them completely coincide. In the example shown in FIG. 4(a), the image part 230a of the bolt hole existing as the mark 230 in the first image frame 220a and the image part 230b of the bolt hole existing as the mark 230 in the second image frame 220b are image parts of the common bolt hole existing on the bottom surface of the vehicle body B. Therefore, the first image frame 220a and the second image frame 220b are combined as image data so that the image part 230a and the image part 230b overlap on the image. By performing such connection using the marks 230 between adjacent image frames for all the image frames 220a to 220z of the bottom surface imaging data 210 (see FIG. 4(b)), a reference image 400 as shown in FIG. 4(c) is generated.

[0047] When generating this reference image 400, the received bottom surface imaging data 210 may be directly decomposed frame by frame and superimposed and synthesized. However, it is desirable to perform grayscale processing on the bottom surface imaging data 210. By performing grayscale processing on the bottom surface imaging data 210, it is possible to make the image clear by removing the color information of unclear parts caused by light reflection, color bleeding, etc., and improve the accuracy of stitching and synthesis. In addition, when the image cannot be sharpened only by grayscale processing, edge extraction processing for extracting outlines or processing for changing the contrast of the image may be performed.

[0048] This reference image 400 is sequentially generated from the image frame 220 that captures the head of the vehicle body B, so that the outer shape of the vehicle body B gradually appears. When the vehicle body B is imaged up to the end, a reference image 400 that includes the head to the end of the vehicle body B in a single still image is generated. The reference image 400 generated by the reference image generation unit 320 is transmitted to the defect position specifying unit 330 of the information processing means 300.

[0049] Next, based on the information transmitted from the defect detection unit 310 and the reference image generation unit 320, the defect position specifying unit 330 specifies the position of the defect.

[0050] The defect position specifying unit 330 determines the corresponding frame 120' from the information about which frame it is from the head of the defect image 120 received from the defect detection unit 310. As shown in FIG. 1, since this corresponding frame 120' arranges the design surface imaging means 100 and the bottom surface imaging means 200 at the same distance from the head of the vehicle body B, the frame number from the head is the same as that of the defect image 120. For example, if the defect image 120 is the 25th frame captured from the head of the vehicle body B, the corresponding frame 120' is also the 25th frame.

[0051] Next, the distance from the head of the vehicle body B to the head of the corresponding frame 120' is calculated in terms of the number of pixels on the screen. Specifically, as shown in FIG. 5(a), the distance from the leading portion of the vehicle body B imaged in the leading image frame 220a to the left end of the corresponding image frame 220 is calculated in terms of the number of pixels (the number of pixels arranged for display on the screen).

[0052] The defect position specifying unit 330 converts the calculated number of pixels into the actual distance in millimeters based on the master information corresponding to the bottom surface imaging means 200 (how many millimeters per pixel). At this time, the distance from the leading end of the calculated vehicle body to the leading end of the corresponding frame 120' is represented by the symbol a in FIG. 5.

[0053] Next, the position of the defective portion 121 shown in the defective image 120 is specified. Specifically, as shown in FIG. 5(b), the distance from the left end (leading end) of the defective image 120 to the defective portion 121 is calculated in terms of the number of pixels (the number of pixels arranged for display on the screen).

[0054] The calculated number of pixels is converted into the actual distance in millimeters based on the master information corresponding to the design surface imaging means 100 (how many millimeters per pixel). At this time, the distance from the left end of the calculated defective image 120 to the defective portion 121 is represented by the symbol b in FIG. 5.

[0055] Finally, by adding the distance from the leading portion of the vehicle body B to the left end of the corresponding image frame 220, which has been converted into the actual distance, and the distance from the leading end of the defective image 120 to the defective portion 121, the distance from the leading end of the vehicle body B to the defective portion 121 can be actually specified (as shown in FIG. 5(c), the distance from the leading end of the vehicle body B to the defective portion is a + b).

[0056] Next, a method for specifying the position of the defective portion 121 in the height direction (vertical direction) will be described. The design surface captured data 110 is based on characteristic points reflected in the captured frame. This reference point shall be a point from which the height can be obtained from information related to manufacturing such as the design drawings of the vehicle. For example, a tire house formed in an arch shape according to the tire shape, etc.

[0057] Then, the height from the reference point to the defective part 121 imaged in the defective image 120 is calculated in terms of the number of pixels (the number of pixels arranged for display on the screen).

[0058] Based on the master information corresponding to the design surface imaging means 100 (how many millimeters per pixel), the calculated number of pixels is converted into the actual height in millimeters from the reference point 122. By adding the height from the converted reference point to the defective part 121 and the height of the reference point obtained from the information related to manufacturing, the height of the defective part 121 in the height direction (vertical direction) can be specified.

[0059] Also, when there is no feature point that can be a reference point in the frame imaged in the design surface captured data 110, the height of the defective part 121 in the height direction (vertical direction) can also be specified from the imaging conditions of the design surface imaging means 100, etc.

[0060] Specifically, since the design surface imaging means 100 always performs imaging at the same angle of view, the height of the upper end or the lower end of the imaging range is measured in advance. When a defect is detected, the distance from the upper end or the lower end of the defective image 120 to the defective part 121 is calculated in terms of the number of pixels (the number of pixels arranged for display on the screen). Based on the master information corresponding to the design surface imaging means 100 (how many millimeters per pixel), the calculated number of pixels is converted into the actual height in millimeters from the reference point.

[0061] By using such a method, even when it is difficult to set the reference point, the height of the defective part 121 in the height direction (vertical direction) can be specified. However, when the reference point can be set, it is preferable to use the method described above because a higher accuracy can be achieved by obtaining the height from the relative relationship within the actually captured frame.

[0062] Also, as a method for specifying the distance from the front end of the vehicle body B to the defective portion 121, the following method can also be used. First, as shown in FIG. 6, when the vehicle body B conveyed by the conveying device C enters the imaging area, the design surface imaging means 100 and the bottom surface imaging means 200 start imaging simultaneously. At this time, as described above, the design surface imaging means 100 and the bottom surface imaging means 200 are arranged so that the distances from the front end of the vehicle body B are the same (see FIG. 1).

[0063] As shown in FIG. 6, the design surface imaging means 100 captures the painted design surface of the vehicle body B, which is the inspection object, and transmits the design surface imaging data 110 to the defect detection unit 310 of the information processing means 300.

[0064] The defect detection unit 310 holds a vehicle body image without the same defect as the vehicle type to be inspected, which is recorded in advance, as comparison imaging data. The defect detection unit 310 that has received the design surface imaging data 110 compares the image of the comparison imaging data with the design surface imaging data 110 to check whether there is a defect in the design surface of the vehicle body B, thereby detecting a defect.

[0065] When a defect is detected, the image frame that captured the defect is recognized as the defect image 120. The design surface imaging data 110 including the defect image 120 is transmitted to the defect position specifying unit 330 of the information processing means 300.

[0066] As shown in FIG. 6, the bottom surface imaging means 200 captures the bottom surface, which is the non-design surface of the vehicle body B, which is the inspection object, and transmits the bottom surface imaging data 210 to the reference image generation unit 320 of the information processing means 300.

[0067] The reference image generation unit 320 of the information processing means 300 decomposes the bottom surface imaging data 210 received from the bottom surface imaging means 200 frame by frame, and combines them to generate a reference image 400 that serves as a reference for identifying the position of defects.

[0068] As shown in FIG. 4, the method for generating the reference image 400 is to use the bolt holes and machining traces reflected in the image frames 220a to 220z obtained by decomposing the bottom surface imaging data 210 frame by frame as marks 230, and overlap and connect the same imaging ranges of the front and rear frames. That is, the mark 230 is an arbitrary image part included in the image frames 220a to 220z and is set when generating the reference image 400. As the mark 230, a special part that is clearly distinguishable from the general part, which is a monotonous image part, is used in each of the image frames 220a to 220z. The mark 230 is, for example, a characteristic image part such as bolt holes and machining traces reflected in each of the image frames 220a to 220z. Specifically, when the mark 230 is a bolt hole, the front and rear image frames 220 are overlapped so that the positions of the bolt holes reflected in them completely coincide. In the example shown in FIG. 4(a), the image part 230a of the bolt hole existing as the mark 230 in the first image frame 220a and the image part 230b of the bolt hole existing as the mark 230 in the second image frame 220b are image parts of the common bolt hole existing on the bottom surface of the vehicle body B. Therefore, the first image frame 220a and the second image frame 220b are combined as image data so that the image part 230a and the image part 230b overlap on the image. By performing such connection using the marks 230 between adjacent image frames for all the image frames 220a to 220z of the bottom surface imaging data 210 (see FIG. 4(b)), a reference image 400 as shown in FIG. 4(c) is generated.

[0069] This reference image 400 is sequentially generated from the image frame 220a that captures the leading end of the vehicle body B, so that the outer shape of the vehicle body B gradually appears. When the vehicle body B is imaged up to the trailing end, a reference image 400 that includes the entire vehicle body B from the leading end to the trailing end in a single still image is generated. The reference image 400 generated by the reference image generation unit 320 is transmitted to the defect position identification unit 330 of the information processing means 300.

[0070] Next, based on the information transmitted from the defect detection unit 310 and the reference image generation unit 320, the defect position identification unit 330 identifies the position of the defect.

[0071] The defect position identification unit 330 calculates the width of the overlapping portion 221 before and after each image frame 220 that forms the reference image 400 received from the reference image generation unit 320 in terms of the number of pixels (the number of pixels arranged for display on the screen).

[0072] Based on the master information corresponding to the bottom surface imaging means 200 (how many millimeters per pixel), the calculated width of the overlapping portion 221 before and after the image frame 220 is converted into an actual distance unit (mm).

[0073] Next, based on the master information corresponding to the design surface imaging means 100 (how many millimeters per pixel), the width of the overlapping portion 221 converted into the distance unit (mm) is converted back into the number of pixels (the number of pixels arranged for display on the screen).

[0074] Since the imaging angles of the design surface imaging means 100 and the bottom surface imaging means 200 are different, the master information (how many millimeters per pixel) naturally has different values. However, since the design surface imaging means 100 and the bottom surface imaging means 200 are arranged so that their positions from the leading end of the vehicle body B are the same, the design surface imaging data 110 can be superimposed and synthesized in the same way as the generation of the reference image 400 by performing the above-described two unit conversions. That is, the width of the overlapping portion 221 of the image frames 220 before and after the reference image 400 can be made to correspond to the design surface imaging data 110, and based on that value, the image frames before and after the design surface imaging data 110 can be overlapped.

[0075] Next, the distance from the leading end of the vehicle body B to the defective portion 121 is calculated by the number of pixels (the number of pixels arranged for display on the screen) on the design surface imaging data 110 on which the overlapping synthesis has been performed.

[0076] Finally, by converting the distance from the leading end of the vehicle body B to the defective portion 121 calculated by the number of pixels (the number of pixels arranged for display on the screen) into an actual distance unit (mm) based on the master information (how many mm per pixel) corresponding to the design surface imaging means 100, the distance from the leading end of the vehicle body B to the defective portion 121 can be specified.

[0077] Also, by obtaining the average value of the widths of the overlapping portions 221 before and after the image frame 220 of the bottom surface imaging data 210, the overlapping synthesis of the design surface imaging data 110 can be performed without generating the reference image 400. However, the speed of the conveying device C is not necessarily constant. Therefore, by generating the reference image 400 and adapting all the widths of the overlapping portions 221 before and after the image frame 220 to the design surface imaging data 110, the accuracy of specifying the position of the defect can be improved even when the conveying speed by the conveying device C is not constant.

[0078] Also, as a method for specifying the distance from the leading end of the vehicle body B to the defective portion 121, the following method can also be used. First, as shown in FIG. 7, when the vehicle body B conveyed by the conveying device C enters the imaging area, the design surface imaging means 100 and the bottom surface imaging means 200 start imaging simultaneously. At this time, the design surface imaging means 100 and the bottom surface imaging means 200 are arranged so that the distances from the leading end of the vehicle body B are the same as described above (see FIG. 1).

[0079] As shown in FIG. 7, the design surface imaging means 100 captures an image of the painted design surface of the vehicle body B which is the object to be inspected, and transmits the design surface imaging data 110 to the defect detection unit 310 of the information processing means 300.

[0080] The defect detection unit 310 holds, as comparison imaging data, a vehicle body video without defects of the same vehicle type as the vehicle type to be inspected recorded in advance. The defect detection unit 310 that has received the design surface imaging data 110 compares the video of the comparison imaging data with the design surface imaging data 110 to check whether there are any defects in the design surface of the vehicle body B, thereby detecting defects.

[0081] When a defect is detected, the image frame capturing the defect is recognized as a defect image 120. The design surface imaging data 110 including the defect image 120 is transmitted to the defect position specifying unit 330 of the information processing means 300.

[0082] As shown in FIG. 7, the bottom surface imaging means 200 captures an image of the bottom surface which is the non-design surface of the vehicle body B which is the object to be inspected, and transmits the bottom surface imaging data 210 to the reference image generation unit 320 of the information processing means 300.

[0083] As shown in FIG. 8, the information processing means 300 decomposes the design surface imaging data 110 received from the design surface imaging means 100 and the bottom surface imaging data 210 received from the bottom surface imaging means 200 frame by frame, and arranges the respective imaging data 110 and 210 in a column.

[0084] Next, the image frame 220 of the bottom surface imaging data 210 corresponding to the number of frames from the first frame capturing the vehicle body B of the defect image 120 is specified, and the distance from the start of the defect image 120 to the defect portion 121 is recognized as a defect line 121' on the bottom surface imaging data 210. At this time, since the design surface imaging means 100 and the bottom surface imaging means 200 are arranged so that the distances from the head of the vehicle body B are the same, by aligning the respective imaging data 110 and 210 in a row, the horizontal distances on the data become the same, and it becomes possible to draw a straight line from the defective portion 121 to the defective line 121'.

[0085] Next, the front and rear image frames 220 of the bottom surface imaging data 210 with the defective line 121' drawn are overlapped based on the landmarks 230 imaged in each image frame 220 to generate a reference image 400.

[0086] Next, the distance from the head of the vehicle body B to the defective line 121' drawn on the reference image 400 is calculated in terms of the number of pixels (the number of pixels arranged for display on the screen).

[0087] Finally, the distance from the head of the vehicle body B to the defective portion 121 calculated in terms of the number of pixels (the number of pixels arranged for display on the screen) is converted into an actual distance unit (mm) based on the master information corresponding to the bottom surface imaging means 200 (how many mm per 1 pixel), whereby the distance from the head of the vehicle body B to the defective portion 121 can be specified.

[0088] As described above, since there are many bolt holes and machining marks on the bottom surface, which is a non-design surface of the vehicle body B, by using them as the landmarks 230, the front and rear of the decomposed image frames 220 can be joined together. Also, by using this joined image as the reference image 400 and calculating the distance from the head of the vehicle body B to the defective portion 121, for example, even when a problem occurs on the production line and the conveying device C stops temporarily, the subsequent image joining can be performed based on the landmarks reflected in the bottom surface imaging data 210, and there is an effect that the detection of the defective position is not hindered.

[0089] Even when the conveyance speed of the vehicle body B by the conveyance device C is not constant, it has no influence on the generation of the reference image 400 or the identification of the position of the defect based on the reference image 400, and the position of the defect generated on the design surface can be identified with high accuracy.

[0090] Moreover, it can be realized by adding the lower bottom surface imaging means 200 and the information processing means 300 for processing the information obtained from the lower bottom surface imaging means 200 to an existing manufacturing line, and can be introduced at low cost.

[0091] [Method for Identifying the Position on the Conveyance Lane of the Vehicle Body] Next, a method for identifying the position of the target vehicle body B on the conveyance lane using the defect position identification device M according to the present invention will be described with reference to the drawings.

[0092] The above-described defect position identification device M can also identify the position of the vehicle body B being conveyed in the front-rear direction of the defect position identification device M on the conveyance lane in a manufacturing line using the conveyance device C.

[0093] The method for identifying the position of the vehicle body B using the defect position identification device M is performed based on the actual movement amount 410 of the vehicle body B per frame by the reference image generation unit 320 that generates the reference image 400 as described in the defect position identification method.

[0094] In order to calculate the movement amount 410, first, the lower bottom surface imaging data 210 imaged by the lower bottom surface imaging means 200 is transmitted to the reference image generation unit 320 of the information processing means 300.

[0095] The reference image generation unit 320 decomposes the lower bottom surface imaging data 210 received from the lower bottom surface imaging means 200 for each frame, and overlaps them to generate the reference image 400.

[0096] As shown in FIG. 4, the method for generating the specific reference image 400 is to use the bolt holes and machining traces reflected in the image frames 220a to 220z obtained by decomposing the bottom surface imaging data 210 frame by frame as the marks 230, and overlap and connect the same imaging ranges of the front and rear frames.

[0097] At this time, the front and rear image frames 220 that are the basis of the reference image 400 always image the bottom surface of the vehicle body B at the same angle of view. Therefore, as shown in FIG. 9, the overlapping portion 221 overlapped when generating the reference image 400 is the range imaged in both the front and rear image frames 220.

[0098] The fact that the overlapping portion 221 is the range imaged in both the front and rear image frames 220 means that for the image frames 220 that are overlapped front and back due to the movement of the vehicle body B, the range from the angle of view or the range newly entered into the angle of view is the actual amount of movement of the vehicle body B advancing per frame. That is, the movement amount 410 is the range of the non-overlapping portion 222 of the image frame 220 when generating the reference image 400. Therefore, by obtaining the width of the non-overlapping portion 222 of the reference image 400, the actual movement amount 410 of the vehicle body B moving per frame can be obtained.

[0099] More specifically, the generation of the reference image 400 for the two front and rear image frames 220a and 220b shown in FIG. 9 will be described as an example. The image frame 220a and the image frame 220b are synthesized so that the same imaging ranges are overlapped based on the marks 230 (bolt holes, machining traces, etc.) imaged on each other.

[0100] At this time, the width of the overlapped range is the overlapping portion 221, and this overlapping portion 221 is the range of the image portion where the imaging objects are common to each other in the image frame 220a and the image frame 220b.

[0101] The non-overlapping part 222 that can be calculated by subtracting the overlapping part 221 from the horizontal width of the image frame 220a is the range that is not imaged in the image frame 220b, which is the next frame of the image frame 220a. In other words, it is the range that has disappeared due to the movement of the vehicle body B within one frame.

[0102] That is, it means that the vehicle body B moves by the width of the non-overlapping part 222 per frame. Calculate the width of this non-overlapping part 222 in terms of the number of pixels (the number of pixels arranged for display on the screen), and based on the master information corresponding to the bottom surface imaging means 200 (how many millimeters per pixel), convert the calculated number of pixels into the actual distance unit (mm), thereby calculating the actual movement amount 410 of the vehicle body B per frame.

[0103] By the way, if the conveyance speed of the vehicle body B conveyed by the conveyance device C is constant, this movement amount 410 will be the same value no matter which image frame 220 of the reference image 400 it is calculated from. However, the conveyance speed of the vehicle body B by the conveyance device C is not necessarily constant. However, in calculating the movement amount 410, the target for position identification is a large object, the vehicle body B moving on the conveyance lane, which is different from the defective part 121 occurring on the above-mentioned design surface. Therefore, as long as the conveyance speed does not fluctuate greatly, the movement amount 410 obtained based on the reference image 400 will have no effect on the position identification of the vehicle body B.

[0104] Next, the position of the vehicle body B on the conveyance lane is identified by multiplying the calculated movement amount 410 by the number of frames after passing through the defective position identification device M.

[0105] Specifically, since the movement amount 410 is the amount by which the vehicle body B actually moves per frame, the value obtained by multiplying the movement amount 410 by the number of frames elapsed since it passed through the defective position identification device M and was no longer imaged is the total movement distance of the vehicle body B after passing through the defective position identification device M. Once the total moving distance is calculated, the position of the vehicle body B on the conveyance lane from the defect position specifying device M can be specified.

[0106] Also, the imaging means such as a video camera has a fixed number of image frames 220 captured per second, and it is also possible to perform calculations by replacing the number of frames with time based on the number of frames captured per second.

[0107] Also, the method for specifying the position of the vehicle body B on the conveyance lane described above can specify the position based on the end point of detection by the defect position specifying device M. Therefore, by providing the reference passing notification means 500 upstream of the conveyance lane from the defect position specifying device M, the position of the vehicle body B located in front of the defect position specifying device M can also be specified.

[0108] Specifically, as shown in FIG. 10, a passing notification means 500 for notifying passing is provided upstream of the defect position specifying device M according to the present invention. The passing notification means 500 may be any means that can determine that the vehicle body B has passed, for example, an imaging means such as a laser sensor or a camera.

[0109] The position where the passing notification means 500 is disposed serves as a calculation reference for specifying the position of the vehicle body B. When the passing notification means 500 confirms that the vehicle body B has passed, it transmits the information to the information processing means 300.

[0110] The vehicle body B that has confirmed passing calculates the number of frames after passing, and multiplies the number of frames by the moving amount 410 calculated by the method described above, thereby specifying the position from the passing notification means 500.

[0111] As described above, the defect position specifying device M according to the present invention specifies the position of the vehicle body B with respect to the defect that has occurred in the vehicle body B. In addition to this effect, it is also possible to specify where the vehicle body B is on the conveyance lane of the conveyance device C.

[0112] The malfunction location identification device M and the malfunction location identification method described above are merely examples, and the configuration of the actual malfunction location identification device M and the malfunction location identification method are not limited thereto. That is, the malfunction location identification device M and the malfunction location identification method according to the present invention are capable of highly accurately identifying the location of a malfunction that has occurred in the vehicle body B, which is the object to be inspected, by having the features described above. The specific malfunction location identification method using the malfunction location identification device M is not limited to the content described above.

[0113] The description of the above-described embodiments is an example of the present invention, and the present invention is not limited to the above-described embodiments. Therefore, even outside the above-described embodiments, various modifications can be made according to the design and the like as long as they do not deviate from the technical idea of the present invention. Also, the various effects described above are merely a list of the preferable effects resulting from the present invention, and the effects of the present invention are not limited to those described in this embodiment.

Explanation of Reference Numerals

[0114] M Location identification device C Conveyor device B Vehicle body 100 Design surface imaging means 110 Design surface imaging data 120 Malfunction image 120′ Corresponding frame 121 Malfunction location 121′ Malfunction line 200 Bottom surface imaging means 210 Bottom surface imaging data 220 Image frame 221 Overlapping part 222 Non - overlapping part 230 Mark 230a Image part 230b Image part 300 Information processing means 310 Malfunction detection unit 320 Reference image generation unit 330 Malfunction location identification unit 400 reference image 410 movement amount 500 passing notification means

Claims

1. A defect position specifying device for specifying the position of defects such as scratches and paint unevenness that occur on the design surface of a vehicle body that is an object to be inspected, design surface imaging means for imaging the design surface of the vehicle body, lower bottom surface imaging means for imaging the lower bottom surface that is the non-design surface of the vehicle body, information processing means for processing information obtained from the design surface imaging means and the lower bottom surface imaging means, and is provided with, the information processing means, a defect detection unit that detects defects occurring on the design surface of the vehicle body based on the design surface imaging data acquired by the design surface imaging means, a reference image generation unit that decomposes the lower bottom surface imaging data acquired by the lower bottom surface imaging means into a plurality of image frames, and generates a reference image for specifying the position of a defect by connecting the front and back of the image frames based on a landmark that is an arbitrary image portion included in the image frames, and a defect position specifying unit that specifies the position of the defect detected by the defect detection unit based on the reference image. A defect position specifying device characterized by the above.

2. A defect position specifying method for specifying the position of defects such as scratches and paint unevenness that occur on the design surface of a vehicle body that is an object to be inspected, acquiring design surface imaging data and lower bottom surface imaging data by imaging the design surface and the lower bottom surface of the vehicle body, performing defect detection on the design surface imaging data based on the acquired information, decomposing the lower bottom surface imaging data into a plurality of image frames, generating a reference image by connecting the front and back of the image frames based on a landmark that is an arbitrary image portion included in the image frames, calculating the position of the defect based on the design surface imaging data and the reference image, converting the distances on the design surface imaging data and the lower bottom surface imaging data into actual distance units to specify the position of the defect. A defect position specifying method characterized by the above.

Images