JP7024604B2 - Body information management system and body information management method - Google Patents

Description

The present invention relates to a technique for managing information regarding inspection results of a vehicle body.

Conventionally, in the painting process of a car body production line, the car body after painting that is continuously transported is inspected, and the inspection result (state of appearance defect, dust adhesion, etc.) is recorded for each car, such as a form. It is known how to manage by handwriting on paper with letters, numbers, symbols and the like. However, such paper information on the recording paper is disadvantageous for comprehensively managing the inspection results for all the vehicle bodies. Therefore, there is a demand for a technique that enables comprehensive management of inspection results for a plurality of vehicle bodies by converting paper information into text data.

Therefore, Patent Document 1 below proposes a production information management device for achieving this kind of technology. This production information management device has a function of extracting characters and the like from an image of paper information on a recording paper and converting it into text data. Therefore, it is effective for comprehensively managing the paper information handwritten on a plurality of recording papers.

Japanese Unexamined Patent Publication No. 2008-257679

By the way, when a recording sheet such as a form is used, the format such as the shape, size, position, and orientation of information that can be entered by handwriting is not easily restricted, and the degree of freedom of handwriting is high. Therefore, in the vehicle body production line, there is a great merit by using the recording paper. On the other hand, the worker needs to complete the entry on the recording sheet within the limited time when the vehicle body is being transported, and even if it is the same worker, it is even more handwritten if it is a different worker. There can be variations in the format of the information provided.

For such a defect, the function of the production information management device disclosed in Patent Document 1 is effective for recognizing the standard information handwritten at a fixed position on the recording paper and converting it into text data. , There is a problem that it is difficult to associate the information randomly handwritten on the recording paper with the pointed out part and convert it into text data.
And, such a problem arises not only in the painting process of the vehicle body production line but also in the case of converting the information handwritten on the recording paper into data in various production processes such as welding process, assembly process and inspection process. obtain.

The present invention has been made in view of the above problems, and is an object of the present invention to provide a vehicle body information management technique capable of associating an inspection result handwritten on a handwritten recording medium with a pointed out portion and converting it into text data. ..

One aspect of the present invention is
It is a car body information management system for managing information on car body inspection results.
An extraction unit that extracts handwritten image data corresponding to the inspection result information from the image data of the handwritten recording medium in which the inspection result information is handwritten with respect to the contour representing the component of the vehicle body.
It is located around the leader line drawn from the contour, the start point and the end point of the leader line, and the end point of the leader line from the inspection result information of the handwritten image data extracted by the extraction unit. The detection unit that detects the quality information and the indication range shown to surround the start point of the leader line .
A conversion unit that converts the quality information detected by the detection unit into text data in which symbol data indicating the position of the start point of the leader line and numerical data indicating the size of the indicated range are linked.
With body information management system,
It is in.

Another aspect of the present invention is
It is a vehicle body information management method that manages information related to vehicle body inspection results by using a vehicle body information management system having an extraction unit, a detection unit, and a conversion unit .
An extraction step of extracting handwritten image data corresponding to the inspection result information from the image data of the handwritten recording medium in which the inspection result information is handwritten with respect to the contour representing the component of the vehicle body by the extraction unit .
It is located around the leader line drawn from the contour, the start point and the end point of the leader line, and the end point of the leader line from the inspection result information of the handwritten image data extracted in the extraction step. A detection step in which the detection unit detects the quality information and the indication range indicated so as to surround the start point of the leader line .
A conversion step of converting the quality information detected in the detection step into text data in which the symbol data indicating the position of the start point of the leader line and the numerical data indicating the size of the indicated range are linked by the conversion unit.
Body information management method, including
It is in.

In each of the above embodiments, handwritten image data corresponding to the inspection result information is extracted from the image data of the handwritten recording medium in which the inspection result information is handwritten with respect to the contour. From this handwritten image data, the starting point coordinates of the leader line and the quality information are detected. Then, the detected quality information is converted into text data associated with the start point coordinates.

Here, the leader line is a line drawn from the outline of the component. Therefore, by detecting the coordinates of the starting point of the starting point, which is the drawing source of the leader line, it is possible to recognize which pointed out point among the constituent elements of the vehicle body the leader line points to. In addition, the product type information is handwritten at the position around the end point, which is the destination of the leader line. Therefore, by detecting this product type information, the specific content of the inspection result can be recognized.

Then, the start point coordinates of the leader line and the product type information are linked to each other to obtain dataized text data. This makes it clear which position of the contour of the component the quality information is, that is, which pointed out point is the inspection result. Therefore, it becomes possible to comprehensively manage the inspection results of a plurality of vehicle bodies based on the text data.

As described above, according to the above aspect, it is possible to provide a vehicle body information management technique capable of associating the inspection result handwritten on the handwritten recording medium with the pointed out portion and converting it into text data.

The system configuration diagram of the vehicle body information management system of Embodiment 1. The figure which shows the image data of the recording paper before the inspection result information is handwritten. The figure which shows the image data of the recording paper which the inspection result information was handwritten. The figure which shows the processing flow in the vehicle body information management system of FIG. The figure for demonstrating the process in the 3rd step of FIG. The figure which shows the handwritten image data extracted in the 4th step of FIG. The figure which shows the judgment information. The figure for demonstrating a part of the process in the 5th step of FIG. The figure for demonstrating a part of the process in the 5th step of FIG. The figure for demonstrating a part of the process in the 5th step of FIG. The figure for demonstrating a part of the process in the 5th step of FIG. The figure for demonstrating a part of the process in the 5th step of FIG. The figure for demonstrating a part of the process in the 5th step of FIG. The figure for demonstrating a part of the process in the 5th step of FIG. The figure which shows the text data which was converted in the 7th step of FIG. The figure for demonstrating a part of the process corresponding to the 5th step of FIG. 4 in the vehicle body information management system of Embodiment 2. FIG.

Preferred embodiments of the above embodiments will be described below.

In the vehicle body information management system, the origin coordinates of the contour in the manual recording medium are stored in advance, and it is preferable that the detection unit detects the start point coordinates of the leader line based on the origin coordinates. ..

According to this management system, the coordinates of the start point of the leader line can be detected based on the relative positional relationship between the contour of the component and the start point of the leader line. Therefore, even if the arrangement of the hull ring in the handwriting recording medium is changed due to a change in the format or the like, the start point coordinates of the leader line can be easily detected by using the origin coordinates as a reference.

In the vehicle body information management system, it is preferable that the detection unit detects the information on the extension line of the leader line as the quality information.

According to this management system, it is possible to accurately detect the quality information handwritten on the extension line as well as the leader line.

In the vehicle body information management system, it is preferable that the detection unit detects information similar to any of a plurality of determination information registered in advance as the quality information.

According to this management system, only information that is highly likely to be quality information can be selectively detected from the information located around the end point of the leader line. Therefore, the quality information detection process becomes simple.

In the vehicle body information management system, the detection unit uses the coordinates of one of the leader line end points located in the inner region surrounded by the contour among the leader line end points on both sides of the leader line in the handwritten image data as the start point. It is preferable to detect it as coordinates.

According to this management system, the end points on both sides of the leader can be easily distinguished into a start point and an end point depending on whether the end points are in the inner region or the outer region of the contour of the component. This makes it possible to improve the detection accuracy of the start point coordinates of the leader line.

In the vehicle body information management system, the extraction unit uses the image data as the first image data and the image data of the manual recording medium before the inspection result information is manually written for the contour as the second image data. When the image data is used, it is preferable to extract the handwritten image data from the first image data as a foreground to the second image data by background subtraction of the second image data from the first image data.

According to this vehicle body information management system, handwritten image data can be selectively extracted by using the process of background subtraction of the second image data from the first image data.

In the vehicle body information management system, the inspection result information is handwritten on the manual recording medium in a color different from the contour, and the extraction unit extracts the handwritten image data from the first image data. In addition, it is preferable to identify and extract this handwritten image data by the difference in color between the handwritten image data and the second image data corresponding to the contour.

According to this vehicle body information management system, it becomes easy to extract handwritten image data from the first image data by operating the inspection result information in a handwritten recording medium so as to be handwritten in a color different from the contour.

In the vehicle body information management system, the detection unit detects the indication range surrounding the start point of the leader line as the inspection result information, and outputs text data indicating the size of the indication range in association with the quality information. It is preferable to do.

According to this vehicle body information management system, the user can grasp the size of the pointed out range based on the output text data.

In the vehicle body information management system, the detection unit uses the text data indicating the size of the indicated range as the text data between the center of gravity of the indicated range and the distance point farthest from the center of gravity in the indicated range. It is preferable to output the range radius corresponding to the maximum distance.

According to this vehicle body information management system, the user can easily grasp the size of the indicated range from the value of the range radius.

The vehicle body information management system preferably includes an image input device for inputting an image of the manual recording medium, and the image input device is preferably connected to the data acquisition unit so as to be able to transmit the image data. ..

According to this vehicle body information management system, image data of a hand-held recording medium can be acquired in cooperation with an image input device.

In the vehicle body information management system, it is preferable that the image input device is arranged at each of a plurality of inspection points in the vehicle body production line.

According to this vehicle body information management system, an operator can input an image of a hand-held recording medium pointed out at each inspection point of the vehicle body production line and handwritten inspection result information by an image input device after each inspection. become.

Hereinafter, the vehicle body information management system of the present embodiment will be described with reference to the drawings.

(Embodiment 1)
As shown in FIG. 1, the vehicle body information management system (hereinafter, also simply referred to as “management system”) 1 according to the first embodiment is for managing information related to the inspection result of the vehicle body 2. In this management system 1, the recording paper 10 is used. The recording paper 10 is a thin, paper-made handwriting recording medium on which the worker U can record information by hand.

In the vehicle body production line, the advantage of using the recording paper 10 having a high degree of freedom in handwriting is great, while the worker U can use the recording paper 10 within the limited time during which the vehicle body 2 is transported. It is necessary to complete the entry in, and even if the same worker U is used, and if the worker U is different, the format of the handwritten information may vary.

Therefore, the management system 1 is for dealing with such a problem, and includes an image input device 20, a processing device 30, an output device 40, a server 50, and a LAN 60.

The image input device 20 is for inputting an image of the recording paper 10 and outputting the digitized image data I of the image. Although the details will be described later, the image data I of the recording paper 10 includes the first image data Ia showing an image in which a plurality of indication information B is handwritten with respect to the contour Wa representing the component W of the vehicle body 2. There is a second image data Ib showing an image before the plurality of pointed out information B is handwritten.

As the image input device 20, a scanner, a camera, or the like can be typically used. The image input device 20 is connected so that the image data I can be transmitted to the data acquisition unit 31 of the processing device 30. At this time, the image input device 20 may be used in a state of being constantly connected to the processing device 30, or is usually used so as to be separated from the processing device 30 and connected to the processing device 30 as needed. May be done. As a result, the processing device 30 can acquire the image data I of the recording paper 10 in cooperation with the image input device 20.

The processing device 30 has a function of processing the image data I sent from the image input device 20. The processing device 30 is composed of a desktop or notebook personal computer (PC) having a monitor (display means), a keyboard (input means), a mouse (selection means), a CPU, a ROM, a RAM, and the like.

The processing device 30 includes a data acquisition unit 31, a data processing unit 32, a storage unit 37, and an input / output unit 38, and the data processing unit 32 further includes an adjustment unit 33, an extraction unit 34, a detection unit 35, and a conversion unit 36. ing.

The data acquisition unit 31 is sent via the image data I sent from the image input device 20, the image data sent via the server 50 and the input / output unit 38, and the LAN 60 and the input / output unit 23. It has a function to acquire image data and the like.

The storage unit 37 includes image data I sent from the image input device 20, image data sent via the server 50 and the input / output unit 38, image data sent via the LAN 60 and the input / output unit 38, and the like. Has a function to store. Further, the storage unit 37 stores in advance information about the origin coordinates (origin coordinates P0 described later) of the contour Wa of the component W on the recording paper 10.

The input / output unit 38 has a function of transmitting / receiving data to / from an external device. This external device is composed of a plurality of OA terminals and the like connected to each of the output device 40, the server 50, and the LAN 60.

The output device 40 is composed of at least one of a printer that outputs data by printing, a monitor that outputs data by display, and a speaker that outputs data by voice.

The server 50 has a function of storing the data input via the LAN 60, reading the stored data in response to the control signal from the request destination, and transmitting the stored data to the request destination.

The adjustment unit 33 of the data processing unit 32 is configured to perform processing for adjusting the size and orientation of the image data I based on the template image data. According to the adjustment unit 33, the two image data Ia and Ib can be adjusted respectively.

Since the two image data Ia and Ib hardly cause a positional deviation with respect to the template image data, the adjustment unit 33 may be omitted when the adjustment processing of the two image data Ia and Ib is not required. ..

The extraction unit 34 of the data processing unit 32 corresponds to the inspection result information B including the leader line D and the quality information C from the first image data Ia by using the two image data Ia and Ib adjusted by the adjustment unit 33. It is configured to perform a process of extracting handwritten image data Ic.

The detection unit 35 of the data processing unit 32 includes quality information C, leader line D, leader line end point E, and indication range F as inspection result information B from the handwritten image data Ic extracted by the extraction unit 34. And are configured to detect each.

The quality information C is information indicating an inspection result handwritten with letters, symbols, numbers and the like. The leader line D is a leader line drawn from the contour Wa of the component W. The leader line end points E are the end points on both sides of the leader line D drawn from the contour Wa of the component W. The pointed out range F is a range surrounding the start point of the leader line D drawn from the contour Wa of the component W.

The conversion unit 36 of the data processing unit 32 links the quality information C detected by the detection unit 35 to the start point coordinates of the leader line (start point coordinates P1, P2, P3, P4 described later) (text data TD described later). ) Is configured to be converted.

Here, each of the above two image data Ia and Ib will be specifically described.

As shown in FIG. 2, in the second image data Ib, four regions A1, A2, A3, and A4 are provided on the paper surface of the recording paper 10.

In the area A1 of the recording paper 10, the contour Wa representing the component W of the vehicle body 2 is described in advance. When the worker U finds the pointed out item during the vehicle body inspection, the inspection result information B as the pointed out item can be recorded by hand on the contour Wa of the area A1.

Here, the component W of the vehicle body 2 may be the vehicle body 2 itself, or may be an element such as a door that is a part of the vehicle body 2. Further, as the contour Wa, those in which the component W is viewed from various directions such as lateral, front, and rear can be used.

The three areas A2, A3, and A4 of the recording paper 10 are arranged side by side above the area A1. Vehicle body management information is handwritten in these three areas A2, A3, and A4. The "vehicle number" is handwritten in the second area A2, the "paint color" is handwritten in the third area A3, and the "date" is handwritten in the fourth area A4.

If necessary, at least one of these three areas A2, A3, and A4 can be diverted to other areas where vehicle body management information is handwritten. Further, the number and arrangement of the areas where the vehicle body management information is handwritten are not limited to this, and can be changed as appropriate.

The four corners 10a printed on the recording paper 10 are used for adjusting the size and orientation of the image data for each of the two image data Ia and Ib in order to match the size and orientation of the image data with the template image data. As this adjustment process, a projection transformation (homography transformation) is typically used.

If necessary, such as when the four corners 10a are not printed on the recording paper 10, four markers for alignment can be embedded in the vicinity of the four corners of the recording paper 10. In this case, the number of markers is not limited to four, and may be changed to more than four if necessary. For example, the recording paper 10 may be divided into a plurality of rectangles, and markers may be embedded in the vicinity of the four corners of each rectangle.

As shown in FIG. 3, in the first image data Ia, the plurality of inspection result information B for the contour Wa of the region A1 of the recording paper 10 includes four quality information C1, C2, C3, C4 (in the following description). , Simply referred to as "quality information C"), four leader lines D1, D2, D3, D4 (also simply referred to as "leader line D" in the following description), and four leader line end points E1, E2. , E3, E4 (also referred to simply as "leader end point E" in the following description) and two indication ranges F1, F3 (also simply referred to as "pointing range F" in the following description). It has been.

As the leader line D, various shapes such as a straight line, a curved line, a wavy line, and a zigzag line can be adopted. The quality information C is located around the end point of the leader line D and is on an extension of the leader line D.

In the case of post-painting inspection of the vehicle body 2, the quality information C is, for example, the characters "Suke" indicating that the paint is too little and the color underneath can be seen through, or the paint is too much and drips. The letters "sauce" indicating the condition, the letters "butsu" indicating the state where solid foreign matter such as dust is on the surface of the paint film, and the surface of the paint film have irregularities such as wrinkles, creases, and folds. It is indicated by the letters "wrinkles" that indicate the condition.

The leader line end point E is the start point and the end point of the leader line D. That is, the leader line end point E1 is the start point and the end point of the leader line D1. Further, the leader line end point E2 is a start point and an end point of the leader line D2. Further, the leader line end point E3 is a start point and an end point of the leader line D3. Further, the leader line end point E4 is a start point and an end point of the leader line D4.

The surrounding portion surrounding the start point of the leader line D for specifying the indication range F is typically composed of a closed shape such as a "circle" or an "ellipse".

In the present embodiment, the indication range F1 is handwritten for the quality information C1, and the indication range F3 is handwritten for the quality information C3. On the other hand, since the quality information C2 and C4 are pointed out pinpointly, the pointed out range is not handwritten.

Next, a vehicle body information management method for managing information by the above management system 1 will be described with reference to FIGS. 1, 4 to 15.

As shown in FIG. 4, this process is executed by the steps from the first step S101 to the eighth step S108. One or more steps may be added to these steps as needed, or a plurality of steps may be integrated. In addition, the order of each step can be changed as needed.

The first step S101 is a step of inputting an image of the recording paper 10 in which the inspection result information B is handwritten to the image input device 20. According to the first step S101, the image of the desired recording paper 10 is input to the image input device 20. The first step S101 may be carried out by the worker U himself who handwritten the inspection result information B on the recording paper 10, or may be carried out by an administrator other than the worker U. Further, the image input device 20 outputs the first image data Ia in a state of being connected to the processing device 30.

Regarding the first step S101, it is preferable that the image input device 20 is arranged at each of a plurality of inspection points in the vehicle body production line. As a result, the worker U can input the image of the recording paper 10 pointed out at each inspection point of the vehicle body production line and handwritten the inspection result information B by the image input device 20 after each inspection.

Further, in this case, it is preferable to carry one recording sheet 10 together with the vehicle body 2 so that the inspection result information B is added by hand to the recording sheet 10 at each inspection location. As a result, as the recording sheet 10 goes around each inspection point, the number of inspection result information B handwritten on the recording sheet 10 increases. Therefore, it is possible to obtain the first image data Ia of the number of recording sheets 10 corresponding to the number of inspection points, which helps to manage the change over time of the pointed out matter with respect to the vehicle body 2.

Of the processing flow of FIG. 4, the processing after the second step S102 is executed by the processing apparatus 30.

The second step S102 is an acquisition step for acquiring image data Ia and Ib output from the image input device 20. This second step S102 is executed by the data acquisition unit 31 of the processing device 30. According to the second step S102, the image data Ia and Ib are taken into the processing device 30. In this case, the acquisition source of the image data Ia and Ib may be the image input device 20, the storage unit 37, or the server 50. The image data Ia and Ib are stored in the storage unit 37 from the data acquisition unit 31 as needed, and then indirectly sent to the adjustment unit 33 of the data processing unit 32, or directly adjusted by the data acquisition unit 31. It is sent to the unit 33.

The third step S103 is an adjustment step for adjusting the shape and orientation of the image data Ia and Ib acquired in the second step S102. This third step S103 is executed by the adjusting unit 33 of the processing device 30. This adjustment is performed for each of the two image data Ia and Ib.

As shown in FIG. 5, for the adjustment of the image data Ia, the coordinates of the four corners 10a included in the image data Ia are detected, and the positions of these four corners 10a are defined by the four corners of the template image data 10b. The shape and orientation of the image data Ia are adjusted to match the position. Since the adjustment of the image data Ib is the same as the adjustment of the image data Ia, the description thereof will be omitted.

The fourth step S104 is an extraction step for extracting the handwritten image data Ic from the first image data Ia. This fourth step S104 is executed by the extraction unit 34 of the processing device 30. In the fourth step S104, the handwritten image data Ic is extracted from the first image data Ia as a foreground to the second image data Ib by background subtraction of the second image data Ib from the first image data Ia.

The term "background subtraction" as used herein refers to the comparison of the pixel values of the first image data Ia and the second image data Ib, and if the degree of difference between the pixel values is equal to or less than the threshold value, it is determined to be the background. It is a known method to determine.

According to the fourth step S104, as shown in FIG. 6, the handwritten image is obtained by omitting the second image data Ib corresponding to the contour Wa (see FIG. 2) of the component W from the first image data Ia. Data Ic is extracted. That is, in the handwritten image data Ic, only the result information B handwritten in the area A1 and the information handwritten in the three areas A2, A3, and A4 remain. In this way, the handwritten image data Ic can be selectively extracted by using the process of background subtracting the second image data Ib from the first image data Ia.

Further, in the fourth step S104, it is preferable that the inspection result information B is handwritten on the recording paper 10 in a color different from the contour Wa of the component W. In this case, when the handwritten image data Ic is extracted from the first image data Ia, the extraction unit 34 identifies the handwritten image data Ic by the difference in color between the handwritten image data Ic and the second image data Ib corresponding to the contour W. .. As a result, by operating the inspection result information B on the recording paper 10 so as to be handwritten in a color different from the contour Wa of the component W, it becomes easy to extract the handwritten image data Ic from the first image data Ia.

In addition, the first image data Ia may include information having low relevance to the quality information C, such as a seal of an inspector or an inspector. In such a case, in order to prevent the detection of the quality information C from being disturbed, the process of deleting the information indicating the seal stamp from the handwritten image data Ic based on the color and the information indicating the seal stamp are handwritten image data based on the shape. It is preferable to perform one of the processes of erasing from Ic. For example, the information indicating the seal can be erased by making the seal black, which is the same color as the border. In addition, information in which characters are surrounded by a shape such as a circle, an ellipse, or a rectangle can be detected and deleted as information indicating a seal.

The fifth step S105 is a detection step of decomposing the handwritten image data Ic and individually detecting each of the quality information C, the leader line D, the leader line end point E, and the indication range F. In the fifth step S105, the quality information C is detected as information on the extension line of the leader line D, so that the quality information C handwritten on the extension line together with the leader line D can be detected with high accuracy. Further, the sixth step S106 is a detection step for detecting vehicle body management information. Both of these two steps S105 and S106 are executed by the detection unit 35 of the processing device 30.

Here, a plurality of determination information is registered in advance in the storage unit 37 for each of the quality information C, the leader line D, the leader line end point E, and the indication range F. Therefore, information similar to any of the plurality of determination information can be selected.

As shown in FIG. 7, the plurality of judgment information JCs regarding the quality information C include images of characters such as "scale", "sauce", "butsu", and "wrinkles", and images showing symbols. It has been. The plurality of determination information JDs regarding the leader line D include images of a plurality of lines having different shapes. The plurality of determination information JEs regarding the leader line end points E include images of a plurality of line end shapes. The plurality of determination information JF regarding the indicated range F includes an image of a surrounding portion (circular, elliptical, etc.) surrounding the line end.

As shown in FIG. 8, in the fifth step S105, the quality information C1, C2, C3, and C4 are detected by the detection unit 35. All of these quality information C1, C2, C3, and C4 are similar to any one of the plurality of determination information JCs. Therefore, the quality information C1, C2, C3, and C4 are determined by comparison with a plurality of determination information JCs. Further, the detection unit 35 detects the pixels in which each of these four quality information C1, C2, C3, and C4 is located as coordinates.

As described above, the detection unit 35 is the information located around the end point of the leader line D and on the extension line of the leader line D, and is similar to any of the plurality of determination information JCs registered in advance. Is detected as quality information C. In this case, since only the information that is highly likely to be quality information can be selectively detected, the detection process of the quality information C becomes simple.

As shown in FIG. 9, in the fifth step S105, the leader lines D1, D2, D3, and D4 are detected by the detection unit 35. All of these leader lines D1, D2, D3, and D4 are similar to any one of the plurality of determination information JDs. Therefore, the leader lines D1, D2, D3, and D4 are determined by comparison with a plurality of determination information JDs.

In the fifth step S105, the leader end points E1, E2, E3, and E4 are detected by the detection unit 35. All of these leader line end points E1, E2, E3, and E4 are similar to any one of the plurality of determination information JEs. Therefore, the leader end points E1, E2, E3, and E4 are determined by comparison with a plurality of determination information JEs. Further, the detection unit 35 detects the pixels in which each of these leader end points E1, E2, E3, and E4 are located as coordinates.

As shown in FIG. 10, since one of the leader line end points E1 on both sides of the leader line D1 is located in the inner region Sa surrounded by the contour Wa, it is recognized as the start point E1a of the leader line D1. To. On the other hand, since the other of the leader line end points E1 on both sides of the leader line D1 is located in the outer region Sb of the contour Wa, it is recognized as the end point E1b of the leader line D1.

Similarly, it is recognized that one of the leader end points E2 on both sides of the leader D2 is the start point E2a of the leader D2 and the other is the end point E2b of the leader D2. Further, it is recognized that one of the leader line end points E3 on both sides of the leader line D3 is the start point E3a of the leader line D3 and the other is the end point E3b of the leader line D3. Further, it is recognized that one of the leader line end points E4 on both sides of the leader line D4 is the start point E4a of the leader line D4 and the other is the end point E4b of the leader line D4.

In this way, the endpoints on both sides of the leader lines D1, D2, D3, and D4 are set to start points E1a, E2a, E3a, and E4a depending on whether the endpoints are in the inner region Sa or the outer region Sb of the contour Wa of the component W. And the end points E1b, E2b, E3b, E4b can be easily distinguished. As a result, the detection accuracy of the start point coordinates P1, P2, P3, P4 of the leader lines D1, D2, D3, D4 can be improved.

Further, the detection unit 35 uses the pixel coordinates of the start points E1a, E2a, F3a, and E4a of the leader lines D1, D2, D3, and D4 as the start point coordinates P1 (x1, y1), P2 (x2, y2), and P3 (x3, y3). ), P4 (x4, y4). On the other hand, the origin coordinates P0 (x0, y0) are set in advance for the contour Wa of the component W, and the information about the origin coordinates P0 is stored in the storage unit 37. The position of each part of the inner region Sa of the contour Wa is specified with reference to the origin coordinate P0. Therefore, by detecting the start point coordinates P1, P2, P3, and P4, it is possible to specify the pointed out points indicated by the start points E1a, E2a, F3a, and E4a.

In this case, the start point E1a of the leader lines D1, D2, D3, D4 based on the relative positional relationship between the contour Wa of the component W and the start points E1a, E2a, F3a, E4a of the leader lines D1, D2, D3, D4. , E2a, F3a, E4a can be detected. Therefore, even if the arrangement of the ring Wa on the recording paper 10 is changed due to a change in the format or the like, the starting points E1a, E2a, F3a, E4a of the leader lines D1, D2, D3, D4 can be used as a reference by using the origin coordinates P0 as a reference. Can be easily detected.

As shown in FIG. 11, in the fifth step S105, the indicated ranges F1 and F3 are detected by the detection unit 35. All of these indication ranges F1 and F3 are similar to any one of the plurality of determination information JFs. Therefore, the indication ranges F1 and F3 are determined by comparison with a plurality of determination information JFs. Further, the detection unit 35 detects the pixel coordinates of the start point E1a surrounded by the indication range F1 as the coordinates of the indication range F1, and the pixel coordinates of the start point E3a surrounded by the indication range F3 are the pixel coordinates of the indication range F3. Detect as coordinates.

Here, as shown in FIG. 12, when the detection unit 35 detects the indication ranges F1 and F3 surrounding the start points of the leader lines D1 and D3 as the inspection result information B, these indication ranges F1 and F3. It is preferable to output the paint color image data Id painted with a predetermined color.

As a result, the worker U outputs the paint color image data Id in the state where the indication ranges F1 and F3 surrounding the start points of the leader lines D1 and D3 are painted to the output device 40 on the screen or prints the quality information C1. , It is easy to visually recognize which pointed out range C3 is intended for.

Further, in the fifth step S105, the detection unit 35 is configured to output the text data T1 and T3, which will be described later, indicating the magnitudes of the indication ranges F1 and F3 in association with the quality information C1 and C3. As a result, the user U can grasp the size of the indication range F1 corresponding to the leader line D1 and the size of the indication range F3 corresponding to the leader line D3 based on the output text data T1 and T3. ..

As shown in FIG. 13, the text data T1 has a range radius r1 corresponding to the maximum distance between the center of gravity G1 of the indication range F1 and the distance point H1 farthest from the center of gravity G1 in the indication range F1. .. This range radius r1 corresponds to the radius of the virtual circle V1 centered on the center of gravity G1. As a result, the user U can easily grasp the size of the indicated range F1 from the value of the range radius r1.

Similarly, as shown in FIG. 14, the text data T3 has a range radius corresponding to the maximum distance between the center of gravity G3 of the indication range F3 and the distance point H3 farthest from the center of gravity G3 in the indication range F3. It is r3. This range radius r3 corresponds to the radius of the virtual circle V3 centered on the center of gravity G3. As a result, the user U can easily grasp the size of the indicated range F3 from the value of the range radius r3.

As the text data T1, the area of the virtual circle V1 may be output instead of or in addition to the range radius r1. Similarly, as the text data T3, the area of the virtual circle V3 may be output in place of or in addition to the range radius r3.

In the sixth step S106, the detection unit 35 detects the information handwritten in each of the three regions A2, A3, and A4 as text data. As a result, text data indicating each of the "vehicle number", "paint color", and "date" can be obtained for the vehicle corresponding to the current recording paper 10.

Here, as the determination information JG for determining the information handwritten in each of the three areas A2, A3, and A4, the storage unit 37 has "0" to "9" as shown in FIG. , The alphabets from "A" to "Z", and the symbols "/" and "#" are registered in advance. Therefore, for example, for the information handwritten in the area A2, the detection unit 35 divides the information in the area A2 into characters one by one, and selects information similar to each character from the determination information JG. As a result, the information handwritten in each of the three areas A2, A3, and A4 can be accurately determined.

The seventh step S107 converts the quality information C1, C2, C3, C4 detected in the fifth step S105 into text data associated with the start point coordinates P1, P2, P3, P4 of the leader lines D1, D2, D3, D4. It is a conversion step to be performed. This seventh step S107 is executed by the conversion unit 36 of the processing device 30. According to the seventh step S107, the text data in which the quality information C1, C2, C3, and C4 are stored in a database is output.

As shown in FIG. 15, the text data TD is a database of information detected in steps S105 and S106 of FIG. 4 in the seventh step S107. As an example, a case where a plurality of combinations of indication information and quality information are described for three vehicle bodies whose vehicle numbers are continuous from T-A001 to T-A003 and whose paint color is the same in CL-01 is shown. There is.

In this text data TD, the symbol "K1" indicating the position of the start point E1a of the leader line D1 (pointed out point Q1) and the value of the range radius r1 of the pointed out point Q1 are added to the character data "scale" indicating the quality information C1. The indicated "a" is associated with. Further, the character data "sauce" indicating the quality information C2 is associated with the symbol "K2" indicating the position of the start point E2a of the leader line D2 (pointed out point Q2). Further, in the character data of "butsu" indicating the quality information C3, the symbol "K3" indicating the position of the start point E3a of the leader line D3 (pointed out point Q3) and the value of the range radius r3 of the pointed out point Q3 "b". And are linked. Further, the character data "wrinkles" indicating the quality information C4 is associated with the symbol "K4" indicating the position of the start point E4a of the leader line D4 (pointed out point Q4). In this way, the inspection result of what kind of indication was made to which part of the vehicle body 2 is converted into text data.

The data output conditions such as the format of the text data TD, the content and the number of items are not limited to those in FIG. 13, and can be appropriately changed as needed. For example, instead of writing the pointed out part by a combination of alphabets and numbers, it is possible to write using kanji, hiragana, katakana, and the like. Further, instead of the form in which the quality information is expressed in katakana, one or a plurality of kanji, hiragana, alphabets, and numbers can be used for display.

The eighth step S108 is a step of storing the text data TD (see FIG. 13) obtained in the seventh step S107. This text data TD is temporarily stored in the storage unit 37, and then sent to the server 50 at any time for storage. The stored text data TD is read from the server 50 as needed, processed into a desired data format, and used for analysis for quality maintenance of the vehicle body and evaluation of the vehicle body production line.

According to the above-mentioned first embodiment, the following effects can be obtained.

In the above management system 1, among the first image data Ia of the recording paper 10 in which the inspection result information B is handwritten with respect to the contour Wa of the component W, the handwritten image data Ic corresponding to the inspection result information B is extracted by the extraction unit 34. Is extracted with. From this handwritten image data Ic, the start point coordinates P1, P2, P3, P4 of the leader lines D1, D2, D3, D4 and the quality information C1, C2, C3, C4 are detected by the detection unit 35. Then, the detected quality information C1, C2, C3, C4 is converted into the text data TD associated with the start point coordinates P1, P2, P3, P4 by the conversion unit 36.

Here, the leader lines D1, D2, D3, and D4 are all lines drawn from the contour Wa of the component W. Therefore, by detecting the start point coordinates P1, P2, P3, P4 of the start point E1a, E2a, E3a, E4a which are the leaders of the leader lines D1, D2, D3, D4, the leader lines P1, P2, P3, It is possible to recognize which pointed out point of the component W of the vehicle body is pointed to by P4. Further, the product information C1, C2, C3, and C4 are handwritten at the peripheral positions of the end points E1b, E2b, E3b, and E4b, which are the destinations of the leader lines D1, D2, D3, and D4. Therefore, by detecting the product type information C1, C2, C3, C4, the specific content of the inspection result can be recognized.

Then, the start point coordinates P1, P2, P3, P4 of the leader lines D1, D2, D3, D4 and the product type information C1, C2, C3, C4 are linked to each other to obtain a text data TD converted into data. As a result, it becomes clear which position of the contour Wa of the component W the quality information C1, C2, C3, C4 is, that is, the inspection result for which pointed out point. Therefore, it becomes possible to comprehensively manage the inspection results of a plurality of vehicle bodies based on the text data TD. In this case, it is effective to keep the man-hours required for data input low as compared with converting the contents handwritten on the recording paper 10 into data by manual input.

As described above, according to the above-described first embodiment, the vehicle body information management technique (vehicle body information management system 1 and vehicle body information management method) capable of associating the inspection result information B handwritten on the recording sheet 10 with the indicated portion and converting it into text data. Can be provided.

Hereinafter, other embodiments related to the above-described first embodiment will be described with reference to the drawings. In other embodiments, the same elements as those in the first embodiment are designated by the same reference numerals, and the description of the same elements will be omitted.

(Embodiment 2)
The management system 1A (see FIG. 1) according to the second embodiment is different from the management system 1 of the first embodiment only in the method in which the detection unit 35 detects each leader line D in the fifth step S105 in FIG. ing.
Other configurations are the same as those in the first embodiment.

As shown in FIG. 16, the detection unit 35 extracts the contour line La of the detection body Da by image processing. After that, the detection unit 35 obtains a quadrangle Lb that approximates the contour line La. Then, the detection unit 35 determines whether or not the detection body Da is the leader line D based on the dimensions of the quadrangle Lb.

In this determination, the detection unit 35 pays attention to the aspect ratio (aspect ratio) of the square Lb, and when the ratio (d2 / d1) of the short side dimension d2 to the long side dimension d1 is lower than the threshold value, the detector Da Can be adopted as a method of determining that is a leader line D, or a method of determining that the detector Da is a leader line D when the dimension d1 of the long side of the square Lb exceeds the threshold value. As a result, the elongated detection body Da is determined as the leader line D, and the detection body Da having any other shape is not determined as the leader line D.

According to the second embodiment, even if there is no information similar to the leader line D in the plurality of determination information JDs (see FIG. 7) of the first embodiment, the leader line D can be detected. Therefore, the determination accuracy of the leader line D can be improved.
Other than that, it has the same effect as that of the first embodiment.

The present invention is not limited to the above-described embodiment, and various applications and modifications can be considered as long as the object of the present invention is not deviated. For example, the following embodiments to which the above-described embodiments are applied can also be implemented.

In the above-described embodiment, the case where each of the quality information C, the leader line D, the leader line end point E, and the indicated range F is individually detected in the fifth step S105 in FIG. 4 is illustrated, but in the reference embodiment, this is illustrated. Instead, only the quality information C and the start point coordinates P1, P2, P3, P4 may be detected as the minimum information necessary for creating the text data.

In the above-described embodiment, the case where the start point coordinates P1, P2, P3, and P4 are detected based on the origin coordinates P0 of the contour Wa of the component W has been illustrated, but instead, it is based on the origin coordinates of the recording paper 10. The origin coordinates P1, P2, P3, and P4 may be detected.

In the above-described embodiment, the case where the information located near the end point of the leader line D and on the extension line of the leader line D is detected as the quality information C has been illustrated. Even if the information is not on the extension line of the leader line D, if it is located near the end point of the leader line D, for example, the information within the specified range from the end point of the leader line D can be detected as the quality information C. ..

In the above-described embodiment, the case where the quality information C is determined by comparison with the determination information JC has been illustrated, but instead of this, the quality information C may be directly detected without using the determination information JC. good.

In the above-described embodiment, the case where a thin and paper-made recording paper 10 is used as the handwriting recording medium has been illustrated, but the thickness, dimensions, materials, and the like of the handwriting recording medium to be used can be appropriately changed.

In the above-described embodiment, the management system 1 effective for managing the inspection result information of the painting process of the vehicle body production line has been exemplified, but this management system 1 is used not only for the painting process of the vehicle body production line but also for the welding process. It can also be used to manage inspection result information in various production processes such as assembly process and inspection process.

1,1A Body information management system (management system)
2 Body 10 Recording paper (handwriting recording medium)
20 Image input device 31 Data acquisition unit 34 Extraction unit 35 Detection unit 36 Conversion unit B Inspection result information C, C1, C2, C3, C4 Quality information D, D1, D2, D3, D4 Leader E, E1, E2, E3 , E4 Leader end points E1a, E2a, E3a, E4a Start point E1b, E2b, E3b, E4b End point F, F1, F2, F3, F4 Pointed out range G1, G3 Center of gravity H1, H3 Separation point I Image data Ia First image data Ib 2nd image data Ic handwritten image data Id paint color image data JC judgment information P0 origin coordinates P1, P2, P3, P4 start point coordinates r1, r3 range radius Sa inner area T1, T3, TD text data W component Wa contour S104 4th step (extraction step)
S105 5th step (detection step)
S107 7th step (conversion step)

Claims (11)

It is a car body information management system for managing information on car body inspection results.
An extraction unit that extracts handwritten image data corresponding to the inspection result information from the image data of the handwritten recording medium in which the inspection result information is handwritten with respect to the contour representing the component of the vehicle body.
It is located around the leader line drawn from the contour, the start point and the end point of the leader line, and the end point of the leader line from the inspection result information of the handwritten image data extracted by the extraction unit. The detection unit that detects the quality information and the indication range shown to surround the start point of the leader line .
A conversion unit that converts the quality information detected by the detection unit into text data in which symbol data indicating the position of the start point of the leader line and numerical data indicating the size of the indicated range are linked.
A vehicle body information management system equipped with. The origin coordinates of the contour in the handwriting recording medium are stored in advance.
The vehicle body information management system according to claim 1, wherein the detection unit detects the start point coordinates of the leader line based on the origin coordinates. The vehicle body information management system according to claim 1 or 2, wherein the detection unit detects information on an extension of the leader line as the quality information. The vehicle body information management system according to any one of claims 1 to 3, wherein the detection unit detects information similar to any of a plurality of determination information registered in advance as the quality information. The detection unit detects the coordinates of one of the leader line end points located in the inner region surrounded by the contour among the leader line end points on both sides of the leader line in the handwritten image data as the start point coordinates. The vehicle body information management system according to any one of 4 to 4. When the extraction unit uses the image data as the first image data and the image data of the manual recording medium before the inspection result information is hand-written for the contour as the second image data, the first image data is used. 1. The item according to any one of claims 1 to 5, wherein the handwritten image data is extracted from the first image data as a foreground to the second image data by background subtraction of the second image data from the image data. Body information management system. The inspection result information is handwritten on the manual recording medium in a color different from the outline.
When the handwritten image data is extracted from the first image data, the extraction unit identifies and extracts the handwritten image data by the difference in color between the handwritten image data and the second image data corresponding to the contour. The vehicle body information management system according to claim 6. The numerical data indicating the magnitude of the indicated range is a range radius corresponding to the maximum distance between the center of gravity of the indicated range and the distance point farthest from the center of gravity in the indicated range, claims 1 to 1. The vehicle body information management system according to any one of 7 . The item according to any one of claims 1 to 8 , further comprising an image input device for inputting an image of the manual recording medium, and the image input device is connected to a data acquisition unit so as to be able to transmit the image data. Described vehicle body information management system. The vehicle body information management system according to claim 9 , wherein the image input device is arranged at each of a plurality of inspection points on the vehicle body production line. It is a vehicle body information management method that manages information related to vehicle body inspection results by using a vehicle body information management system having an extraction unit, a detection unit, and a conversion unit .
An extraction step of extracting handwritten image data corresponding to the inspection result information from the image data of the handwritten recording medium in which the inspection result information is handwritten with respect to the contour representing the component of the vehicle body by the extraction unit .
It is located around the leader line drawn from the contour, the start point and the end point of the leader line, and the end point of the leader line from the inspection result information of the handwritten image data extracted in the extraction step. A detection step in which the detection unit detects the quality information and the indication range indicated so as to surround the start point of the leader line .
A conversion step of converting the quality information detected in the detection step into text data in which the symbol data indicating the position of the start point of the leader line and the numerical data indicating the size of the indicated range are linked by the conversion unit.
Body information management methods, including.

Images