Visual error-proofing system for automobile production line

Abstract

The invention discloses a visual mistake proofing system for an automobile production line, relates to the technical field of industrial automation, and solves the technical problem that in the automobile production process, when the size, the shape and the color of automobile parts possibly do not conform to the standard, the conditions which do not conform to the standard are difficult to classify and process according to a machine visual system. According to the invention, multidirectional part images on an automobile production line are obtained through a monitoring device; acquiring an area number of the automobile part corresponding to the part image; extracting a part image, and identifying the part image to obtain target information of the part; classifying the target information to obtain an operation instruction; and performing corresponding control on the automobile production line according to the operation instruction. According to the method, the size, the shape and the color of the part can be independently analyzed through the obtained omnibearing part image, corresponding operation is carried out on the analysis result, and the quality level of visual mistake proofing of the automobile production line is improved.

Description

Technical Field

[0001] The present invention belongs to the field of industrial automation and relates to computer vision technology, in particular to a visual error-proofing system for an automobile production line. Background Art

[0002] As consumers' demand for personalized and high-quality cars continues to grow, automakers need to produce a variety of configurations and models of cars to meet market demand. This has led to the diversification and complexity of the production process, which has increased the proportion of possible errors in the automobile production process. In order to ensure the correctness of assembly and the quality of the product, automakers need to take effective quality control measures. Traditional automobile production line quality control methods rely on manual inspection and feedback control, but this method is inefficient, difficult to guarantee accuracy, and easily affected by human factors. Therefore, based on computer vision technology, an automobile production line visual error-proofing system has emerged. This system obtains part information through machine vision and analyzes the part information to obtain control instructions for the production line.

[0003] At present, most of the parts images obtained by the visual error-proofing systems of automobile production lines only obtain images of the top, front, back, left and right, and ignore the images in the direction of contact with the conveyor belt. This may lead to incomplete image recognition and, in turn, cause loopholes in the error-proofing system. At the same time, in the process of analyzing part information, when the size, shape, and color of automobile parts do not meet the standards, it is difficult to classify and process the non-standard situations according to the machine vision system, which will reduce the quality level of visual error-proofing of automobile production lines.

[0004] Therefore, the present invention discloses a visual error-proofing system for an automobile production line to solve the above technical problems. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a visual error prevention system for an automobile production line, which is used to solve the technical problem that when automobile parts may not meet the standards in terms of size, shape, or color during automobile production, it is difficult to classify and process the non-standard situations according to a machine vision system. The present invention solves the above-mentioned problem by identifying part images to obtain target information of the parts, classifying the target information to obtain operating instructions, and controlling the automobile production line accordingly according to the operating instructions.

[0006] To achieve the above-mentioned object, the first aspect of the present invention provides a visual error prevention system for an automobile production line, comprising: an error recognition module, and an information collection module, an automatic control module and an information feedback module connected thereto;

[0007] The information acquisition module is used to obtain multi-directional part images on the automobile production line through the monitoring device; obtain the area number where the automobile part corresponding to the part image is located;

[0008] The error recognition module is used to extract the part image, recognize the part image and obtain the target information of the part; classify the target information to obtain the operation instruction; wherein the target information includes the qualified status of size, shape and color;

[0009] The automatic control module is used to obtain operation instructions and control the automobile production line accordingly according to the operation instructions;

[0010] The information feedback module is used to notify the management personnel of the reminder information.

[0011] Preferably, the method of obtaining multi-directional images of parts on the automobile production line through a monitoring device includes:

[0012] The automobile production line is divided into n monitoring areas. A multi-directional camera is installed in the first monitoring area to obtain the initial image of the parts, and a directional camera is installed in the key area to obtain the supplementary image. The initial image and the supplementary image are merged into the part image. Among them, the key areas include the assembly points, quality inspection points, and the areas where the part holes are exposed.

[0013] It should be noted that in the initial image obtained by the multi-directional camera, images of some angles may be difficult to obtain. At this time, the camera in the corresponding direction is needed to obtain supplementary images. The directional camera installed in the present invention is used to obtain supplementary images in a single direction, and the initial image and the supplementary image are merged into a part image, and then the part image is recognized as a whole, which is conducive to obtaining more complete information and providing accurate data for subsequent analysis.

[0014] Preferably, the automobile production line is divided into n monitoring areas, including:

[0015] Get the placement length of the parts and the safety distance between two adjacent parts, and set the sum of the placement length and the safety distance as the standard length of the current part; get the key point area and plan the key area according to the standard length;

[0016] The automobile production line except the key area is divided into several areas according to standard lengths, and the several areas are marked as monitoring areas.

[0017] It should be noted that when planning the key area through the standard length, if the length of the key point area exceeds the standard length, the key point area is divided into m key areas with the position closest to the entrance as the starting point and the standard length as the step length, and the remaining area less than one standard length is also set as a key area; if the length of the key point area does not exceed the standard length, the entire key point area is set as a key area. Planning the key area is conducive to obtaining the area ratio of the corresponding part and area through the standard area size, which is conducive to the subsequent analysis of the size of the part and improves the operating efficiency of the system.

[0018] Preferably, the step of identifying the part image to obtain target information of the part includes:

[0019] Obtain part dimensions through image recognition technology to determine whether the part dimensions meet the standard dimensions; if yes, mark the part dimensions as qualified and proceed to the next step of identification; if no, mark the part dimensions as unqualified and proceed to the next step of identification;

[0020] Get the part shape and determine whether the part shape meets the set shape; if yes, mark the part shape as qualified and proceed to the next step of identification; if no, mark the part shape as unqualified and proceed to the next step of identification;

[0021] Ambient lighting information is obtained through a light measuring device, image lighting information is obtained through image processing technology, and the image is adaptively corrected based on the ambient lighting information and the image lighting information; the corrected part color is obtained through visual recognition technology, and it is determined whether the part color is the specified color; if yes, the part color is marked as qualified; if not, the part color is marked as unqualified; wherein the light measuring device includes a light sensor, a photometer and a spectrometer; the light information includes light intensity and light temperature.

[0022] In the present invention, the part size, part shape, and part color are judged and analyzed separately, which is conducive to obtaining independent information of the part, avoiding mutual interference between different information of the part, and avoiding inaccurate analysis results.

[0023] It should be noted that before judging the color of the parts, correcting the image containing the parts first is conducive to obtaining an image closer to the actual situation and providing more accurate information for subsequent color judgment. Among them, in obtaining image illumination information, the image processing technology can be image histogram statistics, histogram equalization, etc.

[0024] Preferably, the step of classifying the target information to obtain the operation instruction includes:

[0025] Obtain the size of the part marked as unqualified, and determine whether the size of the unqualified part exceeds the fluctuation size; if yes, mark the corresponding part as abnormal situation 1; if no, mark the corresponding part as qualified; wherein the fluctuation size is the fluctuation range set manually;

[0026] Obtain parts with unqualified part shapes, and mark the corresponding parts as abnormal situation 2;

[0027] Get the part whose part color is marked as unqualified, and determine whether the part is a non-appearance part; if yes, set the part color factor YSZ to 0; if no, get the region number i where the part is located, and get the part color factor YSZ through the formula YSZ=exp(a×∑(BGi / i)); where the non-appearance part is manually set; a is the weight adjustment coefficient; the summation range of the summation symbol ∑ is [1,i], and BG is the number of times the part color is unqualified;

[0028] Determine whether the color factor YSZ of the part exceeds the preset threshold; if yes, mark the type corresponding to the part as abnormal situation three; if no, extract the area number i where the part is located, and send the area number i to the management personnel; wherein the preset threshold is obtained through experience.

[0029] The present invention further analyzes the size, shape and color of parts to eliminate possible interference items, thereby increasing the anti-interference capability of the system.

[0030] Preferably, the corresponding control of the automobile production line according to the operation instruction includes:

[0031] Extract abnormal situations. When abnormal situation 1 is received, the area number of the corresponding part is sent to the management personnel, and the subsequent parts are monitored one by one through the monitoring equipment;

[0032] When abnormal situation 2 is received, the area number of the corresponding part is sent to the management personnel, and the automobile production line is stopped;

[0033] When abnormal situation three is received, the area number of the corresponding part is sent to the management personnel and an alarm is issued.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] 1. The present invention obtains target information of the parts by identifying the part images, classifies the target information to obtain operation instructions, and controls the automobile production line accordingly according to the operation instructions. This solves the technical problem that when the size, shape, and color of automobile parts do not meet the standards during the automobile production process, it is difficult to classify and process the non-standard situations according to the machine vision system, thereby improving the quality level of visual error prevention of automobile production lines.

[0036] 2. When acquiring a part image, the present invention first acquires an initial image through a multi-directional camera, and then acquires a supplementary image through a directional camera installed in a key area. The initial image and the supplementary image are combined to obtain an overall part image. The part is identified as a whole through the part image, which is conducive to obtaining more complete information and providing accurate data for subsequent analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0038] Figure 1 It is a schematic diagram of the system module of the present invention;

[0039] Figure 2 It is a schematic diagram of the operation steps of the present invention;

[0040] Figure 3 A schematic diagram of the operation steps for acquiring target information of a part according to the present invention. DETAILED DESCRIPTION

[0041] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] See also Figure 1-Figure 3 , the first aspect of the present invention provides a visual error prevention system for an automobile production line, comprising: an error recognition module, and an information collection module, an automatic control module and an information feedback module connected thereto;

[0043] Information acquisition module: used to obtain multi-directional part images on the automobile production line through the monitoring device; obtain the area number where the automobile part corresponding to the part image is located;

[0044] Error recognition module: used to extract part images, identify the part images to obtain the target information of the parts; classify the target information to obtain operation instructions; the target information includes the qualified status of size, shape, and color;

[0045] Automatic control module: used to obtain operation instructions and control the automobile production line accordingly;

[0046] Information feedback module: used to notify managers of reminder information.

[0047] It should be noted that the area number in the present invention is obtained by installing positioning points on the automobile production line, and the area number moves with the operation of the production line.

[0048] For example: when the automobile production line is stationary, the entire automobile production line conveyor belt is divided into several areas according to the set length, the contact points of adjacent areas are set as positioning points, and the positioning points are numbered in order from the entrance to the exit, and the number of the nearest positioning point in front of the area is set as the number of the current area. When the automobile production line is running, the positioning point will move with the operation of the production line, and the corresponding area number will also move. At this time, if there is a problem with a part in a certain area, the system can report the area number where the part is located to the management personnel, and the management personnel can quickly locate the specific location of the part according to the area number.

[0049] In this application, the monitoring device is used to obtain multi-directional images of parts on the automobile production line, including:

[0050] The automobile production line is divided into n monitoring areas. A multi-directional camera is installed in the first monitoring area to obtain the initial image of the parts, and a directional camera is installed in the key area to obtain the supplementary image. The initial image and the supplementary image are merged into the part image. Among them, the key areas include the assembly points, quality inspection points, and the areas where the part holes are exposed.

[0051] In this embodiment, the monitoring area and the key area are fixed and are responsible for obtaining real-time part images of the corresponding areas. At the same time, the key area is used to obtain images of key positions of parts or positions that are difficult to photograph.

[0052] For example: when the automobile production line is running, images of the parts from top to bottom, front to back, left to right are acquired through the monitoring area. At this time, the image below the parts is blank. When the parts pass through the key area, the directional camera can acquire a supplementary image below the parts by shooting through the shooting holes set below the key area.

[0053] In this application, the automobile production line is divided into n monitoring areas, including:

[0054] Get the placement length of the parts and the safety distance between two adjacent parts, and set the sum of the placement length and the safety distance as the standard length of the current part; get the key point area and plan the key area according to the standard length;

[0055] The automobile production line, except for the key areas, is divided into several areas according to standard lengths, and several areas are marked as monitoring areas.

[0056] It should be noted that when planning the key area through the standard length, if the length of the key point area exceeds the standard length, the key point area is divided into m key areas with the position closest to the entrance as the starting point and the standard length as the step length, and the remaining area less than one standard length is also set as a key area; if the length of the key point area does not exceed the standard length, the entire key point area is set as a key area. Planning the key area is conducive to obtaining the area ratio of the corresponding part and area through the standard area size, which is conducive to the subsequent analysis of the size of the part and improves the operating efficiency of the system.

[0057] It should be noted that m is a positive integer, and the method of comparing the position closest to the entrance is to obtain the closest position by comparing the distance between the position in the key area and the production line of the entrance.

[0058] In this application, the part image is recognized to obtain the target information of the part, including:

[0059] Obtain part dimensions through image recognition technology to determine whether the part dimensions meet the standard dimensions; if yes, mark the part dimensions as qualified and proceed to the next step of identification; if no, mark the part dimensions as unqualified and proceed to the next step of identification;

[0060] Get the part shape and determine whether the part shape meets the set shape; if yes, mark the part shape as qualified and proceed to the next step of identification; if no, mark the part shape as unqualified and proceed to the next step of identification;

[0061] Ambient lighting information is obtained through a light measuring device, image lighting information is obtained through image processing technology, and the image is adaptively corrected based on the ambient lighting information and the image lighting information; the corrected part color is obtained through visual recognition technology, and it is determined whether the part color is the specified color; if yes, the part color is marked as qualified; if not, the part color is marked as unqualified; wherein the light measuring device includes a light sensor, a photometer and a spectrometer; the light information includes light intensity and light temperature.

[0062] In this embodiment, the part size is obtained by image recognition technology to determine whether the part size meets the standard size. The standard size is obtained by comparing the size of the part with the size of the area in the designed area number. The size of the area is used as a reference to obtain the standard size of the part, which can reduce the impact of the distance between the camera and the part on the size of the part in the image, thereby improving the recognition accuracy and efficiency of the system.

[0063] In this application, the target information is classified to obtain operation instructions, including:

[0064] Obtain the size of the part marked as unqualified, and determine whether the size of the unqualified part exceeds the fluctuation size; if yes, mark the corresponding part as abnormal situation 1; if no, mark the corresponding part as qualified; wherein the fluctuation size is the fluctuation range set manually;

[0065] Obtain parts with unqualified part shapes, and mark the corresponding parts as abnormal situation 2;

[0066] Get the parts marked as unqualified in part color, and determine whether the parts are non-appearance parts; if yes, set the part color factor YSZ to 0; if no, get the region number i where the part is located, and get the part color factor YSZ through the formula YSZ = exp(a×∑(BGi / i)); among which, non-appearance parts are manually set; a is the weight adjustment coefficient; the summation range of the summation symbol ∑ is [1, i], and BG is the number of times the part color is unqualified;

[0067] Determine whether the color factor YSZ of the part exceeds the preset threshold; if yes, mark the type corresponding to the part as abnormal situation three; if no, extract the area number i where the part is located and send the area number i to the management personnel; wherein the preset threshold is obtained through experience.

[0068] It should be noted that when judging the size of parts, there may be some parts with low precision requirements. At this time, setting a fluctuation size judgment can make this system suitable for the production of parts with different precisions, thereby improving the scope of application of the system. In the judgment of part color, non-appearance parts are judged. When the parts produced are non-appearance parts, there may not be too high requirements for color. At this time, the system will not strictly restrict the color. When the parts produced are appearance parts, the system will judge the parts based on the preset threshold of the part color factor. When the part color factor exceeds the preset threshold, it proves that the current production line may have a production problem and requires manual identification. When the part color factor does not exceed the preset threshold, the area number where the part is located can be sent to the management personnel for inspection.

[0069] In this application, the automobile production line is controlled accordingly according to the operating instructions, including:

[0070] Extract abnormal situations. When abnormal situation 1 is received, the area number of the corresponding part is sent to the management personnel, and the subsequent parts are monitored one by one through the monitoring equipment;

[0071] When abnormal situation 2 is received, the area number of the corresponding part is sent to the management personnel, and the automobile production line is stopped;

[0072] When abnormal situation three is received, the area number of the corresponding part is sent to the management personnel and an alarm is issued.

[0073] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0074] Working principle of the present invention:

[0075] The present invention obtains multi-directional part images on an automobile production line through a monitoring device; obtains the area number where the automobile part corresponding to the part image is located; obtains the part size through image recognition technology, and determines whether the part size meets the standard size; if yes, marks the part size as qualified and performs the next step of recognition; if no, marks the part size as unqualified and performs the next step of recognition; obtains the part shape, and determines whether the part shape meets the set shape; if yes, marks the part shape as qualified and performs the next step of recognition; if no, marks the part shape as unqualified and performs the next step of recognition; obtains ambient lighting information through a light measuring device, obtains image lighting information through image processing technology, and adaptively corrects the image based on the ambient lighting information and the image lighting information; obtains the corrected part color through visual recognition technology, and determines whether the part color is the specified color; if yes, marks the part color as qualified; if no, marks the part color as unqualified; marks the part size, part shape and part color as target information of the part; classifies the target information to obtain operation instructions; and controls the automobile production line accordingly according to the operation instructions.

[0076] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A visual error-proofing system for automobile production lines, characterized in that: include: An error recognition module, and an information collection module, an automatic control module and an information feedback module connected thereto; The information acquisition module is used to obtain multi-directional part images on the automobile production line through the monitoring device; obtain the area number where the automobile part corresponding to the part image is located; The error recognition module is used to extract the part image, recognize the part image and obtain the target information of the part; classify the target information to obtain the operation instruction; wherein the target information includes the qualified status of size, shape and color; The automatic control module is used to obtain operation instructions and control the automobile production line accordingly according to the operation instructions; The information feedback module is used to notify the management personnel of the reminder information.

2. The automobile production line visual error-proofing system according to claim 1 is characterized in that: The method of obtaining multi-directional images of parts on the automobile production line through the monitoring device includes: The automobile production line is divided into n monitoring areas. A multi-directional camera is installed in the first monitoring area to obtain the initial image of the parts, and a directional camera is installed in the key area to obtain the supplementary image. The initial image and the supplementary image are merged into the part image. Among them, the key areas include the assembly points, quality inspection points, and the areas where the part holes are exposed.

3. The automobile production line visual error-proofing system according to claim 2 is characterized in that: The automobile production line is divided into n monitoring areas, including: Get the placement length of the parts and the safety distance between two adjacent parts, and set the sum of the placement length and the safety distance as the standard length of the current part; get the key point area and plan the key area according to the standard length; The automobile production line except the key area is divided into several areas according to standard lengths, and the several areas are marked as monitoring areas.

4. The automobile production line visual error-proofing system according to claim 1, characterized in that: The step of identifying the part image and obtaining target information of the part includes: Obtain part dimensions through image recognition technology to determine whether the part dimensions meet the standard dimensions; if yes, mark the part dimensions as qualified and proceed to the next step of identification; if no, mark the part dimensions as unqualified and proceed to the next step of identification; Get the part shape and determine whether the part shape meets the set shape; if yes, mark the part shape as qualified and proceed to the next step of identification; if no, mark the part shape as unqualified and proceed to the next step of identification; Ambient lighting information is obtained through a light measuring device, image lighting information is obtained through image processing technology, and the image is adaptively corrected based on the ambient lighting information and the image lighting information; the corrected part color is obtained through visual recognition technology, and it is determined whether the part color is the specified color; if yes, the part color is marked as qualified; if not, the part color is marked as unqualified; wherein the light measuring device includes a light sensor, a photometer and a spectrometer; the light information includes light intensity and light temperature.

5. The automobile production line visual error-proofing system according to claim 4 is characterized in that: The step of classifying the target information to obtain the operation instruction includes: Obtain the size of the part marked as unqualified, and determine whether the size of the unqualified part exceeds the fluctuation size; if yes, mark the corresponding part as abnormal situation 1; if no, mark the corresponding part as qualified; wherein the fluctuation size is the fluctuation range set manually; Obtain parts with unqualified part shapes, and mark the corresponding parts as abnormal situation 2; Get the part whose part color is marked as unqualified, and determine whether the part is a non-appearance part; if yes, set the part color factor YSZ to 0; if no, get the region number i where the part is located, and get the part color factor YSZ through the formula YSZ=exp(a×∑(BGi / i)); where the non-appearance part is manually set; a is the weight adjustment coefficient; the summation range of the summation symbol ∑ is [1,i], and BG is the number of times the part color is unqualified; Determine whether the color factor YSZ of the part exceeds the preset threshold; if yes, mark the type corresponding to the part as abnormal situation three; if no, extract the area number i where the part is located, and send the area number i to the management personnel; wherein the preset threshold is obtained through experience.

6. The automobile production line visual error-proofing system according to claim 5, characterized in that: The corresponding control of the automobile production line according to the operation instructions includes: Extract abnormal situations. When abnormal situation 1 is received, the area number of the corresponding part is sent to the management personnel, and the subsequent parts are monitored one by one through the monitoring equipment; When abnormal situation 2 is received, the area number of the corresponding part is sent to the management personnel, and the automobile production line is stopped; When abnormal situation three is received, the area number of the corresponding part is sent to the management personnel and an alarm is issued.

Images