JP2000067245A - Tire type discriminator - Google Patents

Abstract

(57) [Summary] [PROBLEMS] To improve the efficiency of model pattern extraction work used for tire tread pattern matching, to speed up the matching process, and to improve the accuracy. SOLUTION: A plurality of strip-shaped model patterns are extracted from arbitrary large, medium and small unit patterns randomly arranged on the circumference of a tire. Matching is performed for each strip-shaped model. By changing the interval of the position where the model pattern matching is performed, whether the inspection target image to be matched is any of the large, medium and small unit patterns,
Model patterns can be shared.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for discriminating a type of a vehicle tire, and more particularly to a method for quickly discriminating a tire by a tread pattern and improving work efficiency.

[0002]

2. Description of the Related Art A tire type discriminating apparatus is used in a production line of an automobile to check whether or not a tire according to a production instruction is attached to a vehicle body. By the way, the tread pattern of the tire is 3 in the circumferential direction of the tire.
By randomly arranging up to five basic patterns (unit patterns), road noise such as vibration and noise during traveling is reduced. The plurality of basic patterns are similar patterns having different scales in the circumferential direction.

[0003] When a plurality of similar patterns are included in the tread pattern as described above and are arranged at random, for example, only one of them is provided with a device for determining the tire type based on the tread pattern. Otherwise, there is a high possibility that the apparatus has to check the degree of coincidence with respect to a plurality of imaging positions, which increases the inspection time, which is inconvenient. Therefore, in the conventional apparatus disclosed in Japanese Patent Application Laid-Open No. 9-277806, a plurality of similar patterns are prepared, and the inspection time is reduced by completing the tread pattern recognition processing by one image pickup.

FIG. 4 is an image diagram of processing in a conventional tire type discriminating apparatus. This apparatus obtains a model of a vehicle body to be inspected from production instruction information, extracts a registered model pattern 136 (including similar patterns 138, 140, and 142 corresponding to large, medium, and small pitches) corresponding to the model, and outputs the model data to the vehicle body. An image of a tread pattern in a range corresponding to the model pattern is cut out as an inspection pattern 146 from an image 144 obtained by capturing the mounted inspection target tire. Then, a matching process is performed between the inspection pattern 146 and the registered model pattern 136 to obtain a degree of coincidence.

FIG. 5 is a flowchart for explaining a procedure for obtaining the degree of coincidence from the inspection pattern 146 and the registered model pattern 136. A normalized correlation coefficient is obtained between the grayscale data value of each pixel of the inspection pattern 146 and the grayscale data value of each pixel of each of the patterns 138, 140, and 142 (step 150). This normalized correlation coefficient is obtained for each of the three model patterns, and is set as the coincidence R (correspondence 152, 154, 156) for each pattern. The maximum value of these three coincidences R is calculated (step 15).
8) The maximum value is set as the degree of coincidence 160 of the tread pattern of the inspection pattern 146 for the tire type specified in the specification of the inspection target vehicle type. When the degree of coincidence 160 exceeds a predetermined threshold value, it is determined that the tread patterns match.

In the normalized correlation processing, brightness data at points at the same position in the registered pattern and the pattern to be inspected are respectively sampled to form one set of brightness data. A set of brightness data for a plurality of sampling points is plotted, for example, taking the brightness of the inspection pattern on the vertical axis and the brightness of the registered pattern on the horizontal axis. As the degree of coincidence between the two patterns becomes higher, the brightness data set is distributed along a straight line having a positive slope, and the correlation coefficient of the normalized correlation for the set of the brightness data sets becomes closer to 1.

[0007]

In the above-mentioned conventional method, a plurality of unit patterns having different scales in the circumferential direction are extracted from the tread patterns of the tires used for production and registered in the apparatus in advance. Had to be kept. At that time, since the plurality of unit patterns may be randomly arranged on the circumference of the tire, the work of identifying, extracting, and registering each of the plurality of similar unit patterns within the entire range of the tread pattern requires time and effort. There was a problem. Conventionally, the unit pattern to be matched is
Because of the relatively large size, firstly, for example, a gradation of lightness is generated on a curved tire surface depending on how the light hits, or a local change in lightness due to adhesion of dirt on the surface causes the coincidence by the normalized correlation. Second, there is a problem that the accuracy of the degree determination may be reduced. Second, there is a problem that the processing time of the matching for one unit pattern is relatively long. Further, since matching has to be performed for each unit pattern, there is a problem that the calculation load of the apparatus increases.

[0008] The present invention solves the above-mentioned problems, improves the speed and efficiency of the operation of registering information on unit patterns of registered tires in the device, and improves the processing speed and accuracy of the tire type discrimination processing. It is an object to provide a discriminating device.

[0009]

According to the present invention, there is provided a tire type discriminating apparatus, comprising: a matching pattern for holding a plurality of strip patterns extracted from a plurality of strip-shaped regions arranged in a circumferential direction in a unit pattern of a registered tire; It is characterized by having a holding means, and a pattern comparing means for comparing the tread pattern and the strip pattern of the tire to be inspected sequentially with respect to each of the strip patterns, thereby performing a comparison process of the unit pattern.

According to the present invention, a plurality of strip-shaped areas are defined in a unit pattern, and a tread pattern included in each of the strip-shaped areas is extracted as a strip pattern.
The strip-shaped regions are arranged in the circumferential direction in parallel with each other, for example, with the longitudinal direction thereof oriented in the width direction of the tire. The strip pattern is held in the matching pattern holding means of the apparatus, and using the strip pattern, the pattern comparing means determines whether the tread pattern of the tire to be inspected includes the unit pattern of the registered tire. The pattern comparison means
The first strip pattern is selected, and a location corresponding to the first strip pattern is searched for on the tread pattern of the tire to be inspected. When it is found, the second strip pattern is selected.
The position of the second strip pattern on the tire to be inspected can be narrowed down based on the detection position of the first strip pattern. For example, the search for the second strip pattern is performed within that range. Similarly, the search for the n-th strip pattern is performed when the detection of the first to (n-1) -th strip patterns selected so far is successful.

In the tire type discriminating apparatus according to the present invention, the tread pattern includes a plurality of unit patterns which are similar patterns having different scales in the circumferential direction.
The matching pattern holding unit holds a set of the strip patterns extracted from one unit pattern,
The pattern comparison means performs the comparison process on the similar pattern using a set of strip patterns extracted from the one unit pattern.

According to the present invention, even when the tread pattern includes a plurality of unit patterns having different scales in the circumferential direction as the unit patterns, the matching pattern holding means is extracted from any one of them. It is sufficient to hold a set of strip patterns. The pattern comparing means can detect a unit pattern other than the unit pattern from which the strip pattern is extracted on the inspection target tire by changing the interval between the strip patterns according to the difference in the circumferential scale between the unit patterns. it can.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a schematic configuration diagram of a mounted tire type inspection system to which a tire type identification device of the present invention is applied in an automobile production line. This system confirms that the tires mounted on the vehicle body in the vehicle production line are as specified, and determines the tire type based on the tread pattern.

The vehicle 2 with the tires is moved along a line by the transport system 4. Sensors LS1 and LS2 for detecting the position of the vehicle are provided on the line. Based on the detection results of the sensors LS1 and LS2, a tire type determination process is started. Specifically, when the sensor LS1 operates, the type determination processing for the front wheels of the vehicle is executed, and when the sensor LS2 operates, the type determination processing for the rear wheels of the vehicle is executed.

The main body of the system for inspecting the type of mounted tires includes a CCD camera 6, a driving device 8 for changing the direction of the camera, and a controller 1 for controlling the operation of the driving device 8.
0, a production instruction capturing device 12, a color extracting device 14 for extracting desired color information from an image signal from the CCD camera 6, an image processing device 16, and a programmable device for controlling these components based on information from various components. A configuration including a line controller (PLC) 18;

The CCD camera 6 captures a color image of the tread surface of the tire 20. The CCD camera 6 is arranged, for example, on the floor of a line through which the vehicle body passes so as to easily photograph the tires 20 attached to the vehicle body. CCD
The photographing direction of the camera 6 can be optimally adjusted by the driving device 8, and for example, the elevation angle of the photographing direction is adjusted between the photographing of the front wheels and the photographing of the rear wheels. The drive device 8 is controlled by the controller 10 based on an instruction from the PLC 18.

A video signal from the CCD camera 6 is input to a color extraction device 14. There is a process of performing tire type determination based on color information included in a tire image. For example, it is used to determine the tire type based on tread identification lines such as red, green, and yellow drawn in the circumferential direction of the tread surface of the tire. In such a case, the color extraction device 1
4 extracts video information corresponding to the intensity of a specific color from the video signal from the CCD camera 6. The present invention is a tire type discriminating process based on a tread pattern. Since a lightness (luminance) signal based on a mixture of each color is basically used, a signal corresponding to each color is output from the color extracting device 14. Or a brightness signal based on their combination is output.

The image processing device 16 performs the above-described tire type determination processing based on the tread identification line based on the video signal obtained by the CCD camera 6, or performs the tire type determination processing using the tread pattern according to the present invention.

In the present apparatus, information on a tread pattern relating to a tire that may appear on a production line is stored in a storage unit such as a memory of the image processing apparatus 16 in advance.
Then, when the PLC 18 obtains the identification number of the tire to be attached to the vehicle body transported on the line from the production instruction capturing device 12 and passes the identification number to the image processing device 16, the image processing device 16 A matching process is performed using a pattern corresponding to the identification number.

FIG. 2 is a schematic diagram for explaining the information of the tread pattern stored in the image processing device 16. As described above, a plurality of unit patterns are arranged in the circumferential direction of the tire, and the unit patterns are similar patterns having different scales in the circumferential direction. 2A to 2C show unit patterns whose scale in the circumferential direction is large, medium, and small, respectively. In the figures (a) to (c), the horizontal direction corresponds to the width direction of the tire, the vertical direction corresponds to the circumferential direction,
The solid lines are the unit patterns 30, 32, and 34, respectively. In this apparatus, the unit pattern of the large scale “large”
A set of strip patterns as tread patterns in one of strip-shaped areas (areas elongated in the width direction of the tire surrounded by dotted lines in the figure) 36-1, 36-2, and 36-3; Strip-shaped areas 38-1, 38-2, 38-3
A set of strip patterns that are tread patterns in the inside, or three strip-shaped areas 40-1, 4 of unit patterns of small scale “small”
Only one of a set of strip patterns as tread patterns in 0-2 and 40-3 is extracted from the tread pattern and registered in the image processing device 16. These strip-shaped regions have a common shape regardless of the size of the unit pattern.

That is, in the present apparatus, there is no need to extract strip patterns from all of the large, medium, and small unit patterns and register them in the image processing apparatus 16. The reason for this is that the unit patterns 30, 32, and 34 are similar to each other with different scales in the circumferential direction of the tire. For example, in the figure, the upper strip-shaped regions 36-1, 38 in each of the large, medium, and small unit patterns.
This is because a high correlation coefficient can be obtained by observing the normalized correlation between the strip patterns extracted from -1 and 40-1. Therefore, the strip patterns extracted from any one of the strip-shaped areas 36-1, 38-1, and 40-1 can be shared. For example, using the strip pattern extracted from the strip shape area 38-1 of the "medium" unit pattern, the "large" and "small" strip shape areas 36-1 and 40-1 are used.
, A high degree of matching is obtained. This is the middle strip-shaped area 36 of each of the large, medium and small unit patterns.
-2, 38-2, 40-2 and the strip pattern extracted from any of them, and also from the lower strip-shaped areas 36-3, 38-3, 40-3 and any of them. The same applies to the matching with the strip pattern to be extracted.

Strictly speaking, for example, the upper strip-shaped areas 36-1, 38-1, and 40-1 in each of the large, medium, and small unit patterns have the same strip width (periphery) regardless of the difference in the circumferential scale. Direction), the patterns contained therein are slightly different. However, here, since the area sampled from each unit pattern is long in the width direction of the tire and has a strip shape narrow in the circumferential direction, the pattern extracted from them is reduced in the influence of the difference in the scale in the circumferential direction. Thus, the commonality of the features of the pattern in the width direction of the tire is emphasized. Therefore, the strip patterns have a high degree of coincidence with each other, and one of them can be used for the matching process for each unit pattern. For example, in the present apparatus, model 1, model 2, and model 3, which are strip patterns extracted from the upper, middle, and lower strip-shaped regions 38-1, 38-2, and 38-3 of the "medium" unit pattern, It is registered in the image processing device 16 for each tire type.

FIG. 3 is a schematic flowchart of the tire type discriminating process of the present apparatus. An image of the tread surface of the tire 20 is captured using the CCD camera 6 (S5).
0). This image signal is passed to the image processing device 16 via the color extraction device 14 as a brightness signal. The image processing device 16 receives the production instruction from the production instruction capturing device 12 via the PLC 18, and acquires the tire number specified therein. Then, three model patterns of the tire type corresponding to the tire number are loaded into the matching processing unit (not shown) of the image processing device 16 (S5).
2). Specifically, the model pattern is a set of brightness values defined for each sampling point. The matching processing unit is specifically configured in the present apparatus as a program executed by a CPU built in the image processing apparatus 16,
The data of each sampling point forming the model pattern is loaded into, for example, an array area defined in the program.

The image processing device 16 cuts out a predetermined range including the unit pattern from the image obtained from the CCD camera 6 as an image to be inspected, and starts a matching process.
First, a matching process with the model 1 is performed (S5).
4). In this matching process, a position where the degree of coincidence R is maximum is searched while shifting the positional relationship between the cut-out inspection target image and the model 1 in the tire circumferential direction by, for example, a fixed amount. Here, since the model is a strip pattern which is a part of the unit pattern, the number of sampling points included in the model is smaller than the sampling points of the entire unit pattern used in the conventional matching. Therefore, the time required for the matching process using the strip pattern model is reduced as compared with the matching for the entire unit pattern.

If the maximum coincidence R obtained in the processing S54 is smaller than a predetermined threshold, a rejection is determined (S56). On the other hand, when the maximum matching degree R with respect to the model 1 is equal to or larger than the predetermined threshold, a matching process with the model 2 is performed next (S58). In this model 2 matching, the position of the middle strip pattern in the inspection target image is estimated for each of the large, medium and small unit patterns based on the position where the model 1 matches, and the model 2 matching You can narrow the range. That is, in the present apparatus, the range in which matching with the model 2 is performed and the search for the position where the degree of matching is maximized is narrowed compared to the entire inspection target image, and the processing time is shortened compared to the matching processing of the model 1.

If the maximum coincidence R obtained in step S58 is less than the predetermined threshold value, rejection is determined, while if the maximum coincidence R for the model 2 is equal to or greater than the predetermined threshold value. Then, the process proceeds to a matching process with the model 3 (S60).

In the matching process S62 for the model 3, the position where the model 3 is to be matched is estimated based on the interval between the positions where the model 1 and the model 2 match, and the model is placed in an area having a predetermined upper and lower width centered on that position. The matching range of No. 3 is narrowed down. The matching range is determined according to the accuracy of the estimated value of the matching position of the model 3, and can be determined to be narrower than in the matching process of the model 2. Therefore, the processing time of the matching process for the model 3 is further reduced than that of the model 2.

When the maximum coincidence R obtained in the process S62 is less than the predetermined threshold value, the rejection is determined similarly to the models 1 and 2. On the other hand, if the maximum matching degree R with respect to the model 3 is equal to or larger than the predetermined threshold value, the matching is successful, and it is determined that the tires are attached to the vehicle body according to the production instruction (S64).

Here, the matching of this apparatus is model 1,
This is performed for each of 2 and 3, and the normalized correlation is also examined individually. For this reason, even when the brightness locally changes in the inspection target image, for example, when gradation occurs, the influence of the gradation on each strip-shaped region is reduced, and the degree of coincidence can be accurately evaluated. That is, for example, even if the tire from which the model is extracted and the tire of the image to be inspected are of the same type, when gradation occurs, the correlation coefficient decreases in the matching of the entire unit pattern, and May be determined. On the other hand, in the present apparatus, the normalized correlation is separately performed for each strip region, and the same effect as the level adjustment is performed for each region is obtained. .

As described above, in the present apparatus, a model pattern is extracted from a plurality of strip-shaped regions in one arbitrary unit pattern on a tread pattern of a tire. Then, by changing the interval between the strip patterns and matching the inspection target image, the model pattern can be shared with each of the plurality of similar patterns. Thereby, the efficiency of the model pattern registration work can be improved. In addition, by performing matching for each of a plurality of small regions in the unit pattern, accuracy can be improved and processing time can be reduced.

The strip-shaped region is preferably taken out from the center in the width direction of the tire. The width of the strip varies depending on the tire, but is set to, for example, about 5 pixels of the CCD camera 6.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of a mounted tire type inspection system to which a tire type identification device of the present invention is applied in an automobile production line.

FIG. 2 is a schematic diagram illustrating tread pattern information stored in the image processing apparatus according to the embodiment of the present invention.

FIG. 3 is a schematic processing flowchart of a tire type determination process according to the embodiment of the present invention.

FIG. 4 is an image diagram of a process in a conventional tire type determination device.

FIG. 5 is a flowchart illustrating a procedure for obtaining a degree of coincidence from an inspection pattern and a registered model pattern in the related art.

[Explanation of symbols]

2 car, 4 transport system, 6 CCD camera, 1
2 Production instruction capture device, 16 image processing device, 20 tires, 30, 32, 34 unit pattern.

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2F065 AA51 AA61 BB05 CC13 DD06 FF01 FF04 FF67 JJ03 JJ07 JJ26 PP01 QQ25 QQ31 QQ38 RR08 5B057 AA01 BA02 CA01 CA08 CA12 CA16 CC03 CE09 CE11 DA03 DB02 DB06 DC09 DC32

Claims (2)

[Claims] 1. A tire type discriminating apparatus for discriminating a tire type of an inspection target tire by comparing unit patterns constituting a tread pattern repeatedly arranged in a circumferential direction of the tire with the inspection target tire and a registered tire. A matching pattern holding unit that holds a plurality of strip patterns extracted from each of the plurality of strip-shaped regions arranged in the circumferential direction in the unit pattern of the tire; and the tread pattern and the strip pattern of the inspection target tire. A pattern comparing unit that performs a matching process on the unit patterns by sequentially performing matching on the strip patterns. 2. The tire type discriminating apparatus according to claim 1, wherein the tread pattern includes a plurality of unit patterns which are similar patterns having different scales in the circumferential direction. A set of the strip patterns extracted from the unit pattern is held, and the pattern comparing unit performs the comparison process on the similar pattern using the set of strip patterns extracted from the one unit pattern. A tire type discriminating apparatus, characterized in that: