Crankshaft surface defect detection device based on machine vision

Abstract

The invention discloses a crankshaft surface defect detection device based on machine vision, and relates to the technical field of crankshaft surface defect detection devices. Through cooperation of a first servo motor, a second bolt rod and a movable plate, multi-angle shooting of the detection module is facilitated, the defect recognition precision is improved, and then the function of precisely recognizing tiny defects can be achieved; an acquisition module is used for acquiring images according to the minimum defect size, so that no missing detection of fine defects is ensured, the detection comprehensiveness is improved, invalid data acquisition is reduced, and the detection efficiency is improved; an analysis module adopts a mode of firstly performing mean value pre-screening and then performing accurate comparison on abnormal blocks during gray value analysis, so that the one-to-one comparison work amount of image blocks is greatly reduced, the detection time of a single crankshaft is remarkably shortened, and the method can adapt to the takt time requirement of a production line; the closed contour direction is formed based on the gray value deviation direction and the coordinate connecting line, gray abnormality is judged, the defect recognition rate is increased, and missing detection and misjudgment are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of crankshaft surface defect detection devices, in particular to a crankshaft surface defect detection device based on machine vision. BACKGROUND

[0002] The crankshaft is the core component of the internal combustion engine to realize energy conversion, and the surface quality directly affects the running accuracy, service life and reliability of the engine. Cracks, scratches, pits and other surface defects (the minimum size of the defects can reach 0.1mmx0.1mm) are easily generated in the forging and processing of the crankshaft. If these defects are not detected and removed in time, stress concentration may occur during engine operation, which may cause serious failures such as crankshaft fracture (the failure rate is more than 5%). However, the existing crankshaft surface defect detection device based on machine vision has low detection efficiency in use. The detection time of a single crankshaft is more than 3 minutes, which cannot adapt to the production line with a beat of 60 seconds / root. In addition, the defect recognition rate is only 75%-80% due to the influence of the experience and fatigue of the detection personnel, and the missed detection rate of fine cracks (depth less than 0.05mm) is as high as 20%, which seriously threatens the product quality. In addition, although some automatic detection devices use machine vision technology, the moving flexibility of the detection mechanism is insufficient, and most of them are fixed angle shooting. The complex curved surface parts such as the journal and the fillet of the crankshaft are easy to form a detection blind area (the blind area coverage rate is more than 15%), which leads to a defect misjudgment rate of more than 10%. Therefore, the above technical problems need to be solved and processed. SUMMARY

[0003] The purpose of the present application is to solve the problems existing in the prior art, and a crankshaft surface defect detection device based on machine vision is provided.

[0004] In order to achieve the above purpose, the present application adopts the following technical scheme: a crankshaft surface defect detection device based on machine vision, comprising a device shell and a storage opening formed on one side of the device shell. A cavity is formed in the inner side of the device shell, and a supporting plate is horizontally and fixedly connected and installed in the middle of the inner side of the device shell. A cross-moving mechanism is installed inside the device shell above the supporting plate, and a moving spraying mechanism is installed on the supporting plate. A positioning mechanism is assembled on one side of the moving spraying mechanism. The control box of the detection device is provided with an intelligent control assembly, which comprises a collection module, an analysis module and an execution module. The collection module collects the image data of the crankshaft rotation and transmits the collected image data to the analysis module. The analysis module receives the image data transmitted by the acquisition module, analyzes the length of the portion extending outside the crankshaft rotating shaft, and sets the image acquisition time interval; analyzes the concentrated portion of the image data collected in the corresponding time period and compares it with the grayscale value of the standard part to determine the cause of the grayscale abnormality; establishes a coordinate system based on the number of the grayscale abnormal image block, draws a graph within the coordinate system, and determines the cause of the grayscale abnormality based on the aspect ratio of the image; combines the results of the two grayscale abnormality cause determinations to determine the cause of the grayscale abnormality, generates a warning signal based on the cause, and transmits the warning signal to the execution module; The execution module receives the signal transmitted by the analysis module and performs corresponding operations.

[0005] Preferably, the analysis module performs the following steps to analyze the collection time interval: S1: Connect the centers of the clamping positions at both ends , with the connecting line as the rotation axis, the detection module collects the image data of the crankshaft rotation process at a set time, and connects the line again on the collected image Draw and select any part on the crankshaft image to connect the line Draw a perpendicular line and record its length; S2: The position corresponding to the maximum vertical line length is the farthest distance point; the distance data between the farthest distance point and the corresponding connecting line on multiple collected images is obtained, and the one with the largest distance data value is selected as the selected value ; S3: Angular velocity data of crankshaft rotation To obtain, select the value Angular velocity of the corresponding point , unit is mm / s; the minimum size of the crankshaft surface defect can reach 0.1mm×0.1mm, so in order to ensure that the collected image can accurately reflect the defects on the crankshaft surface, the collection time interval must ensure the acquisition of the minimum size image, that is, the collection time interval .

[0006] Preferably, the analysis module performs the following steps to analyze the grayscale value: K1: Grayscale processing is performed on multiple image data of standard parts collected at the same time, and they are segmented according to the size of the pixel blocks. The image blocks formed after segmentation are numbered according to the number of rows and columns they are in. The total number of rows of image data is divided by two to obtain the number of the middle row. Then, the height and ratio of the pixel block are compared with the size of 0.1mm to determine the row number corresponding to the size of 0.1mm. , with the middle row and the sides of the middle row To record images; K2: Acquire and record the gray value data of each image block in the recorded image, and calculate the mean value and standard deviation , to calculate the mean value and standard deviation , and set the fluctuation range , and the detected gray value outside the fluctuation range is an abnormal value, and after removing the abnormal value, the mean value of the remaining gray value data is calculated, and the mean value is taken as the gray value data of the corresponding numbered gray block of the standard part; K3: Calculate the mean value of the gray values of all image blocks in the corresponding row of the corresponding number, and then calculate the mean value of the gray values of all image blocks in the corresponding row of the standard part, if , it is determined that there is no gray abnormality in the corresponding row; otherwise, the gray values of all image blocks in the corresponding row are compared with the image blocks in the corresponding column of the corresponding row of the standard part; K4: Acquire the gray value data of the image block of the corresponding number in the corresponding row and the gray value data of the image block of the corresponding number in the corresponding row of the standard part, if is not within the range and , it is determined that the gray abnormality is caused by protrusion or crack; if is not within the range and , it is determined that the gray abnormality is caused by scratch or pit.

[0007] Preferably, the analysis module performs the aspect ratio analysis step as follows: Q1: Mark the image block with gray value not within the range as an abnormal image block, after completing the comparison of all image blocks, acquire the number of abnormal image blocks, and establish a coordinate system with the row and column numbers of the image block as the horizontal and vertical coordinates, draw the coordinate points corresponding to the number of abnormal image blocks in the coordinate system, and connect the adjacent coordinate points; Q2: Measure the length data and width data of the closed contour formed by the connecting line, if , it is determined that the gray abnormality is a scratch or crack; otherwise, it is determined that the gray abnormality is a pit or protrusion.

[0008] Preferably, the cross moving mechanism comprises a moving frame fixedly connected to the upper end of the inner part of the device shell, the moving frame is provided with guide grooves horizontally around, and moving screws are crossly and slidingly installed in the guide grooves, one end of the moving screws is coaxially fixedly connected with the output end of the driving motor, and the moving screws are sleeved on the moving seat.

[0009] Preferably, the mechanical arm is fixedly installed at the lower end of the moving seat, and the detection module is fixedly connected to the other end of the mechanical arm.

[0010] Preferably, the moving grooves are oppositely formed in the upper end of the supporting plate, and the recess is upwardly formed in the lower end of the supporting plate.

[0011] Preferably, the moving spraying mechanism comprises a first bolt rod rotatingly installed in the recess, a second servo motor is installed on the device shell at one end of the first bolt rod, the output end of the second servo motor is coaxially fixedly connected with one end of the first bolt rod through the device shell, a spraying cylinder is slidingly installed on the upper end of the supporting plate and the first bolt rod, and a plurality of spraying heads are equidistantly installed in the spraying cylinder.

[0012] Preferably, the positioning mechanism comprises a fixed plate installed on the upper end of the supporting plate, a protective cover is installed on the upper end of the fixed plate, a positioning plate is rotatingly installed at the lower end of the protective cover, a limiting rod is fixedly connected to the other end of the positioning plate, the limiting rod is slidingly connected with the fixed plate, an extension spring is sleeved on the limiting rod between the fixed plate and the positioning plate, a moving plate is installed at the relative position of the fixed plate, the moving plate is slidingly installed in the moving groove, a positioning plate is rotatingly installed on the upper end of the moving plate, a fixed box is installed on the moving plate at the other end of the positioning plate, a rotating motor is installed in the fixed box, the output end of the rotating motor is coaxially fixedly connected with one end of the positioning plate through the moving plate, a second bolt rod is horizontally rotatingly installed at the bottom of the supporting plate at the lower end of the moving plate, a first servo motor is installed on the device shell at one end of the second bolt rod, and the output end of the first servo motor is coaxially fixedly connected with one end of the second bolt rod through the device shell.

[0013] Compared with the prior art, the present application has the following beneficial effects: 1. The moving frame of the cross moving mechanism cooperates with the moving screw rod and the moving seat, which facilitates comprehensive coverage of the surface of the crankshaft, reduces the detection blind area, improves the comprehensiveness of detection, and thus realizes the function of no dead angle detection; the first servo motor of the positioning mechanism cooperates with the second screw rod and the moving plate, which facilitates multi-angle shooting of the detection module and improves the accuracy of defect identification, and thus realizes the function of accurate identification of subtle defects; the detection module cooperates with the cross moving mechanism and the positioning mechanism, which facilitates the detection module to quickly collect the image of the surface of the crankshaft and analyze it, cooperates with the high-efficiency operation of each mechanism, greatly shortens the detection time, improves the detection efficiency, and thus realizes the function of rapid detection of the crankshaft, finally solves the problems of low efficiency, low recognition rate and existence of detection blind area of the existing crankshaft surface defect detection device based on machine vision; 2. The acquisition module collects images according to the minimum defect size, ensures that there is no missed detection of subtle defects, improves the comprehensiveness of detection, reduces invalid data acquisition, and improves the detection efficiency; the analysis module uses the method of "preliminary mean pre-screening and then accurate comparison of abnormal blocks" when analyzing the gray value, which greatly reduces the engineering quantity of image block-by-block comparison, significantly shortens the detection time of a single crankshaft, and can adapt to the production line beat demand; the direction based on the gray value deviation and the direction based on the coordinate line to form a closed contour direction to determine the gray abnormality, improve the defect recognition rate, and reduce the missed detection and misjudgment. BRIEF DESCRIPTION OF DRAWINGS

[0014] The drawings described herein are used to provide further understanding of the present application, constitute a part of this application, the illustrative embodiments of the present application and the description thereof are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings: Figure 1 The overall three-dimensional structure schematic diagram of the present application is shown in the figure; Figure 2 Another side of the overall three-dimensional structure schematic diagram of the present application is shown in the figure; Figure 3 The overall three-dimensional structure schematic diagram of the present application is shown in the figure; Figure 4 The internal overall three-dimensional structure schematic diagram of the present application is shown in the figure; Figure 5 The detection mechanism overall three-dimensional structure schematic diagram of the present application is shown in the figure; Figure 6 The front view cross-sectional structure schematic diagram of the present application is shown in the figure; Figure 7 The Figure 6 The structure of A part in the figure is enlarged; Figure 8 The detection mechanism and spraying mechanism overall three-dimensional structure schematic diagram of the present application is shown in the figure; Figure 9The detection mechanism and the spraying mechanism are shown in the overall perspective view from the bottom. Figure 10 The system flowchart is shown in the figure.

[0015] In the figure, the serial numbers are as follows: 1, device shell; 2, first servo motor; 3, storage opening; 4, second servo motor; 5, supporting plate; 6, moving groove; 7, spraying cylinder; 8, moving frame; 9, moving seat; 10, mechanical arm; 11, detection module; 12, extension spring; 13, protective cover; 14, moving plate; 15, first bolt rod; 16, second bolt rod. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments.

[0017] Embodiment 1: see Figures 1-9 The device shell 1 inside is provided with a cavity, and a supporting plate 5 is horizontally and fixedly connected and installed in the middle of the inside of the device shell 1. A cross moving mechanism is installed inside the device shell 1 above the supporting plate 5. A moving spraying mechanism is installed on the supporting plate 5. A positioning mechanism is assembled on one side of the moving spraying mechanism. The device shell 1 and the supporting plate 5 build a basic frame. The cross moving mechanism, the moving spraying mechanism and the positioning mechanism cooperate to facilitate the detection and processing of the surface defects of the crankshaft. The cross moving mechanism includes a moving frame 8 fixedly connected to the inside upper end of the device shell 1. Guiding grooves are horizontally provided around the moving frame 8. Moving screws are crosswise and slidingly installed in the guiding grooves. The moving screws are coaxially and fixedly connected to the output ends of drive motors at one end. The moving screws are all sleeved on moving seats 9. The moving frame 8, the guiding grooves, the moving screws and the drive motors facilitate the cross direction movement of the detection components. The moving seats 9 are fixedly installed at the lower end of the mechanical arm 10. The other end of the mechanical arm 10 is fixedly connected with a detection module 11. The mechanical arm 10 at the lower end of the moving seat 9 and the detection module 11 facilitate the accurate detection of the surface defects of the crankshaft. The moving grooves 6 are oppositely provided at the upper end of the supporting plate 5. The recesses are upwardly provided at the lower end of the supporting plate 5. The moving grooves 6 on the supporting plate 5 and the recesses at the lower end facilitate the installation of the moving spraying mechanism and provide moving space for the moving spraying mechanism.

[0018] In the application, the mobile spraying mechanism comprises a first bolt rod 15 rotatably installed in a groove, a second servo motor 4 installed on a device housing 1 at one end of the first bolt rod 15, the output end of the second servo motor 4 penetrating the device housing 1 and fixedly connected with the one end of the first bolt rod 15 in a coaxial manner, a spraying cylinder 7 slidably installed on the upper end of a supporting plate 5 with the first bolt rod 15, and a plurality of spraying heads equidistantly installed inside the spraying cylinder 7, so that the detected defects can be marked conveniently through the first bolt rod 15, the second servo motor 4, the spraying cylinder 7 and the spraying heads inside the spraying cylinder 7.

[0019] Embodiment 2: see Figure 10 The control box of the detection device is internally provided with an intelligent control assembly, which comprises a collection module, an analysis module and an execution module. The collection module collects image data of the rotation of the crankshaft and transmits the collected image data to the analysis module. The analysis module receives the image data transmitted by the collection module, analyzes the length of the extension part outside the rotation shaft of the crankshaft, sets the image collection time interval, analyzes the concentrated part of the image data collected in the corresponding time period, compares it with the gray value of the standard part, determines the reason for the gray abnormality, establishes a coordinate system according to the number of the gray abnormal image block, draws a graph in the coordinate system, determines the reason for the gray abnormality according to the length-width ratio of the image, combines the results of the two times of gray abnormality determination, obtains the reason for the gray abnormality, generates a warning signal according to the reason, and transmits the warning signal to the execution module. After the crankshaft is clamped and fixed at both ends, the centers of the clamping positions at both ends are connected , with the connecting line as the rotation axis, the detection module 11 collects the image data of the crankshaft rotation process at a set time, and connects the line again on the collected image Draw and select any part on the crankshaft image to connect the line Draw a perpendicular line and record the length of the perpendicular line; the position corresponding to the maximum perpendicular line length is the farthest distance point; obtain the distance data between the farthest distance point and the corresponding connecting line on multiple collected images, and select the one with the largest distance data value as the selected value ; Angular velocity data of crankshaft rotation To obtain, select the value Angular velocity of the corresponding point , unit is mm / s; the minimum size of the crankshaft surface defect can reach 0.1mm×0.1mm, so in order to ensure that the collected image can accurately reflect the defects on the crankshaft surface, the collection time interval must ensure the acquisition of the minimum size image, that is, the collection time interval ; The standard part is rotated and the image is captured; multiple image data of the standard part captured at the same time are gray-scaled and segmented according to the size of the pixel blocks. The image blocks formed after segmentation are numbered according to the number of rows and columns they are in; when numbering the rows, the row where the corresponding point on the crankshaft is located is the first row, and the subsequent rows are arranged in sequence; the total number of rows of image data is divided by two to obtain the number of the middle row, and then the height and ratio of the pixel block are compared with the size of 0.1mm to determine the row number corresponding to the size of 0.1mm , with the middle row and the sides of the middle row To record the image; obtain and record the gray value data of each image block in the recorded image, and perform the average and standard deviation Calculation of the mean value and standard deviation Set the fluctuation range , for the detected gray values ​​that are not within the fluctuation range, the outliers are removed and the remaining gray value data are averaged. Calculation of the mean value Gray value data of the gray blocks corresponding to the numbers of the standard parts; Average the grayscale values ​​of all image blocks in the corresponding row of the corresponding number Then calculate the mean gray value of all image blocks in the corresponding row of the standard part ,like , it is determined that there is no grayscale abnormality in the corresponding row; otherwise, the grayscale values ​​of all image blocks in the corresponding row are compared with the image blocks with the corresponding row and column numbers of the standard parts; the grayscale value data of the image blocks with the corresponding numbers in the corresponding rows are obtained. Gray value data of the image block with the corresponding number in the corresponding row of the standard part ,like Not present Within the range and , then the grayscale anomaly is determined to be caused by a bulge or crack; the bulge will concentrate the reflection of light to form a highlight area, resulting in a high grayscale value; the sharp angles on both sides of the crack may produce specular reflection of light, causing a sudden increase in the local grayscale value; if Not present Within the range and , the grayscale abnormality is determined to be caused by scratches or pits. When the surface is scratched, the grooves will scatter light, and the light reflected to the detection module will be reduced, resulting in a low grayscale value. The light at the bottom of the pit is easily absorbed, and the reflected light is small, and the grayscale value is lower than the standard range. The value of is obtained through calibration of a large number of samples; analysis of 500 defective samples found that the grayscale fluctuation of normal surfaces is usually ≤8% (due to roughness and lighting), while defects (such as cracks and scratches) can cause grayscale deviations of more than 15% (the reflection at the crack edge is enhanced by 20%+, and the reflection at the scratch groove is reduced by 30%-); therefore, 0.1 can effectively distinguish between "normal fluctuation" and "defect abnormality"; The grayscale value of the corresponding numbered image block is not The marks within the range are abnormal image blocks. After completing the comparison of all image blocks, the numbers of the abnormal image blocks are obtained, and a coordinate system is established with the row and column numbers of the image block numbers as the horizontal and vertical coordinates. The coordinate points corresponding to the abnormal image block numbers are drawn in the coordinate system, and adjacent coordinate points are connected. Then, the length data of the closed contour formed by the connection is calculated. and width data Take measurements, if , the grayscale anomaly is determined to be a scratch or crack; otherwise, the grayscale anomaly is determined to be a pit or bulge; coefficient The function of the metric is to distinguish between long strip and nearly circular defects. Its setting is based on the geometric characteristics of common defects: cracks and scratches are mostly long strips (length is much greater than width), while pits and protrusions are mostly nearly circular (length and width are similar). Through statistics of 500 typical defects, it is found that the aspect ratio of scratches / cracks is usually greater than 3, while the aspect ratio of pits / protrusions is ≤2.5. is a reasonable threshold - when It is judged as a scratch / crack when it is smaller than 0.1mm, otherwise it is a pit / bump. This value needs to be adapted to the defect size: for a minor defect of 0.1mm, It can be appropriately lowered (e.g. 2.5) to avoid misjudgment due to incomplete contours; In combination with the comparison result of the gray scale range, the cause of the abnormal gray scale can be determined, a warning signal is generated, and the warning signal is transmitted to the execution module; After receiving the warning signal, the execution module controls the buzzer module of the intelligent control assembly to issue a buzzer warning, and transmits the cause of the abnormal gray scale to the staff in the form of a short message through the wireless transmission module of the intelligent control assembly, so that the staff can centrally process the corresponding defective crankshaft.

[0020] Working principle: when the present application is used, first, the crankshaft to be detected is placed on the supporting plate 5 through the placing opening 3, so that the two ends of the crankshaft are located between the positioning plates of the fixed plate and the moving plate 14, then the first servo motor 2 is started, the second bolt rod 16 is driven to rotate through the output end, then the moving plate 14 slides along the moving groove 6 through the threaded connection with the second bolt rod 16, until the positioning plate on the moving plate 14 approaches one end of the crankshaft, the positioning plate on the fixed plate cooperates with the positioning plate on the moving plate 14 under the elastic force of the extension spring 12, so as to clamp and fix the crankshaft from both ends to prevent the crankshaft from shaking during detection, then during detection, the second servo motor 4 is started, the first bolt rod 15 is driven to rotate through the output end, so that the spraying cylinder 7 moves along the supporting plate 5, then the marking paint is sprayed out through the spraying head in the spraying cylinder 7 to mark the surface of the crankshaft, then the surfaces of the crankshaft at different angles need to be detected, the rotary motor in the fixed box on the moving plate 14 is started, the positioning plate is driven to rotate through the output end, and then the crankshaft is rotated to the appropriate angle, so as to facilitate the detection module 11 to scan comprehensively, then the drive motor of the cross-moving mechanism is started, the moving screw is rotated in the guide groove of the moving frame 8, so as to drive the moving seat 9 to move crossly along the X and Y axes, so that the mechanical arm 10 and the detection module 11 can cover the entire surface of the crankshaft.

[0021] The above is only the preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can make equivalent replacement or change according to the technical scheme and the inventive concept of the present application within the technical range disclosed by the present application, which should be covered in the protection scope of the present application.

Claims

1. A machine vision-based device for detecting surface defects of a crankshaft, comprising a device housing (1) and a storage opening (3) formed on one side of the device housing (1), characterized in that: The device shell (1) is internally provided with a cavity, and a supporting plate (5) is horizontally fixedly connected to the middle of the inner side of the device shell (1), a cross moving mechanism is internally installed above the supporting plate (5), and a moving spraying mechanism is installed on the supporting plate (5), and a positioning mechanism is assembled on one side of the moving spraying mechanism; The control box of the detection device is internally provided with an intelligent control assembly, which comprises a collection module, an analysis module and an execution module; The collection module collects image data of the rotation of the crankshaft and transmits the collected image data to the analysis module; The analysis module receives the image data transmitted by the collection module, analyzes the length of the extension part outside the rotation shaft of the crankshaft, sets the image collection time interval, analyzes the concentrated part of the image data collected in the corresponding time period, compares it with the gray value of the standard part, judges the reason for the gray abnormality, establishes a coordinate system according to the number of the gray abnormal image block, draws a graph in the coordinate system, judges the reason for the gray abnormality according to the length-width ratio of the image, combines the results of the two times of judgment of the reason for the gray abnormality, obtains the reason for the gray abnormality, generates a warning signal according to the reason, and transmits the warning signal to the execution module; The execution module receives the signal transmitted by the analysis module and performs corresponding operation.

2. The machine vision based surface defect detection apparatus for crankshaft as claimed in claim 1 wherein: The analysis module analyzes the collection time interval as follows: S1: connecting the center of the clamping position at both ends With the connecting line as the rotation axis, the detection module (11) collects image data during the rotation of the crankshaft at a set time, and the connecting line is drawn again on the collected image , and an arbitrary part of the crankshaft image is selected to draw a vertical line , and record the length of the vertical line; S2: take the position corresponding to the maximum perpendicular length as the farthest distance point; acquire distance data between the farthest distance points on the plurality of collected images and the corresponding connecting lines, and select the maximum distance data value as the selected value ; S3: angular velocity data of the crankshaft rotation If the acquisition is performed, the value is selected The rotation angular velocity of the corresponding point Unit: mm / s; the minimum size of the surface defect of the crankshaft can reach 0.1mm*0.1mm, so as to ensure that the collected image can accurately reflect the surface defect of the crankshaft, the collection time interval must ensure that the image collection of the minimum size is realized, that is, the collection time interval .

3. The machine vision based surface defect detection apparatus for crankshaft as claimed in claim 1 wherein: The analysis module analyzes the gray value as follows: K1: multiple image data collected at the same time to the standard piece is gray processing, and according to the size of the pixel block segmentation, the image block formed after the segmentation is numbered according to the row and column; the total number of image data is divided by two to get the number of the middle row, then according to the height size and proportion of the pixel block and the size 0.1mm, the corresponding row number of the size 0.1mm is determined , the middle row and the two sides of the middle row are recorded as images; K2: Acquire and record the gray value data of each image block in the recorded image, and calculate the mean value and standard deviation , and set the fluctuation range , and set the fluctuation range , and set the fluctuation range , and set the fluctuation range , and set the fluctuation range , and set the fluctuation range K3: Calculate the mean value of the gray scale of all image blocks corresponding to the line in the number , and then calculate the mean value of the gray scale of all image blocks corresponding to the line of the standard part , if , it is determined that there is no gray scale anomaly in the corresponding line; Conversely, the image blocks of the corresponding row are compared with the image blocks of the corresponding column number of the standard part in terms of gray value; K4: obtaining the gray value data of the corresponding numbered image block corresponding to the line and the standard part , if is not in the range of and , it is determined that the gray abnormality is caused by protrusion or crack; if is not in the range of and , it is determined that the gray abnormality is caused by scratch or pit.

4. The machine vision based surface defect detection apparatus for crankshaft as claimed in claim 1 wherein: The analysis module analyzes the length-width ratio as follows: Q1: The grayscale value of the corresponding numbered image block is not The image blocks within the range are marked as abnormal. After completing the comparison of all image blocks, the numbers of the abnormal image blocks are obtained. A coordinate system is established with the row and column numbers of the image block numbers as the horizontal and vertical coordinates. The coordinate points corresponding to the abnormal image block numbers are plotted in the coordinate system, and adjacent coordinate points are connected. Q2: Length data of the closed contour formed by the connecting lines and width data are measured, and if then the gray scale anomaly is determined to be a scratch or a crack; otherwise, the gray scale anomaly is determined to be a dent or a bump.

5. The machine vision based surface defect detection apparatus for crankshaft as claimed in claim 1 wherein: The cross moving mechanism comprises a moving frame (8) fixedly connected to the upper end of the inner side of the device shell (1), guide grooves are horizontally formed around the moving frame (8), and moving screws are crosswise and slidingly installed in the guide grooves, one end of each moving screw is coaxially fixedly connected to the output end of a driving motor, and each moving screw is sleeved on a moving seat (9).

6. The machine vision based surface defect detection apparatus for crankshaft as claimed in claim 5 wherein: A mechanical arm (10) is fixedly installed at the lower end of the moving seat (9), and the other end of the mechanical arm (10) is fixedly connected to a detection module (11).

7. The machine vision based surface defect detection apparatus for crank shaft as claimed in claim 1 wherein: A moving groove (6) is oppositely formed in the middle of the upper end of the supporting plate (5), and a recess is upwardly formed in the lower end of the supporting plate (5).

8. The machine vision based surface defect detection apparatus for crankshaft as claimed in claim 7 wherein: The moving spraying mechanism comprises a first bolt rod (15) rotationally installed in the recess, a second servo motor (4) is installed on the device shell (1) at one end of the first bolt rod (15), the output end of the second servo motor (4) is coaxially fixedly connected to one end of the first bolt rod (15) through the device shell (1), a spraying cylinder (7) is slidingly installed on the upper end of the supporting plate (5) and the first bolt rod (15), and a plurality of spraying heads are equidistantly installed in the interior of the spraying cylinder (7).

9. The machine vision based surface defect detection apparatus for crankshaft as claimed in claim 8 wherein: The positioning mechanism comprises a fixed plate installed on the upper end of the supporting plate (5), a protective cover (13) installed on the upper end of the fixed plate, a positioning plate rotatably installed on the lower end of the protective cover (13), a limiting rod fixedly connected to the other end of the positioning plate, the limiting rod being slidably connected to the fixed plate, a telescopic spring (12) sleeved on the limiting rod between the fixed plate and the positioning plate, a moving plate (14) installed in the relative position of the fixed plate, the moving plate (14) being slidably installed in the moving groove (6), a positioning plate rotatably installed on the upper end of the moving plate (14), a fixed box installed on the moving plate (14) at the other end of the positioning plate, a rotary motor installed in the fixed box, the output end of the rotary motor penetrating through the moving plate (14) and being coaxially fixedly connected to one end of the positioning plate, a second bolt rod (16) horizontally rotatably installed on the bottom of the supporting plate (5) at the lower end of the moving plate (14), a first servo motor (2) installed on the device shell (1) at one end of the second bolt rod (16), and the output end of the first servo motor (2) penetrating through the device shell (1) and being coaxially fixedly connected to one end of the second bolt rod (16).

Images