Methods, devices, classification systems, and storage media for elbow screening.

Abstract

This invention discloses a method, apparatus, classification system, storage medium, and computer program for screening elbows. It includes a vision system comprising a first camera and a second camera. The first camera captures a side view image of the elbow, and the second camera captures a bottom view image of the elbow. The method includes determining measurement data of the elbow based on the side and bottom view images, and determining whether the elbow is a qualified or defective product based on the measurement data. This solution, by using a vision system to calculate the actual measurement data of the elbow and determining whether the elbow is a qualified product based on the actual measurement data, achieves automatic classification and screening of elbows, improves the efficiency of automated air conditioning production, prevents defective elbows from being installed in air conditioners, and ensures high product quality in air conditioners.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of air conditioner manufacturing, and particularly relates to a bend screening method and device, a bend screening and classifying system, a storage medium and a computer program product. BACKGROUND

[0002] The two-wire device generally refers to the evaporator and the condenser in the air conditioning system, which are two core components in the air conditioning refrigeration cycle. In the automatic production process of the two-wire device, pipe installation (i.e., two-wire device insertion of the bend) is an extremely important link, which is seriously affected by the incoming bend. The reason is that the incoming bend has various shapes and sizes in the industrial production site pipeline, and various defects may exist in the insertion process of the bend, especially the misplacement and missing of the bend pipe opening, which may cause poor tightness. This not only causes the feeding mechanism of the automatic bend insertion machine to be blocked, but also more seriously reduces the sealing of the pipe connection, causes the refrigerant in the refrigerant system to leak and corrode the electrical components, and directly affects the production quality of the air conditioner. Therefore, for the situation that the bend has various shapes and sizes, manual visual recognition is usually used. However, this method is not only low in efficiency and prone to errors, but also high in labor cost.

[0003] The above content is only used to assist in understanding the technical solutions of the application and does not represent the acknowledgement of the above content as prior art. SUMMARY

[0004] The application aims to provide a bend screening method, device, bend screening and classifying system, storage medium and computer program product, to solve the problem that in the automatic production process of the two-wire device of the air conditioner, the accuracy of the bend insertion process is reduced and the sealing quality of the two-wire device pipe is seriously reduced due to the various shapes and sizes of the bend, so as to calculate the actual measurement data of the bend by setting a visual system, determine whether the bend is a qualified product according to the actual measurement data, realize the automatic classification and screening of the bend, improve the efficiency of the automatic production of the air conditioner, prevent the installation of the defective bend on the air conditioner, and ensure the high product quality of the air conditioner.

[0005] The application provides a bend screening method, which is provided with a visual system, the visual system comprising: a first camera and a second camera; the first camera is used for collecting a side view image of the bend; the second camera is used for collecting a top view image of the bend; the method comprises: acquiring the side view image of the bend collected by the first camera and the top view image of the bend collected by the second camera; determining measurement data of the bend according to the side view image and the top view image; and determining whether the bend is a qualified product or a defective product according to the measurement data.

[0006] In some embodiments, the measurement data of the elbow includes a height of the elbow, a girth welding height of the elbow, a nozzle diameter of the elbow, and a nozzle center distance of the elbow; and determining the measurement data of the elbow according to the side view image and the bottom view image includes: determining the height of the elbow and the girth welding height of the elbow according to the side view image, and determining the nozzle diameter of the elbow and the nozzle center distance of the elbow according to the bottom view image.

[0007] In some embodiments, determining the height of the elbow and the girth welding height of the elbow according to the side view image includes: rectifying distortion of the side view image; locating features and positions of the elbow in the side view image, and correcting positions of the elbow by using preset correction parameters; framing a bottom ROI region of the elbow to determine pixel coordinates of a reference straight line of the bottom of the elbow; framing a top ROI region of the elbow to determine pixel coordinates of a top vertex of the elbow; respectively framing a left girth welding ROI region and a right girth welding ROI region of the elbow to determine pixel coordinates of a first point and a second point of the left girth welding, and pixel coordinates of a first point and a second point of the right girth welding; inputting the pixel coordinates of the top vertex of the elbow and the pixel coordinates of the reference straight line of the bottom of the elbow into a preset elbow height calculation formula to obtain pixel coordinates of the height of the elbow; inputting the pixel coordinates of the first point and the second point of the left girth welding, the pixel coordinates of the first point and the second point of the right girth welding, and the pixel coordinates of the reference straight line of the bottom of the elbow into a preset girth welding height calculation formula to obtain pixel coordinates of the girth welding height of the elbow; and respectively converting the pixel coordinates of the height of the elbow and the pixel coordinates of the girth welding height of the elbow into world coordinates to obtain the height of the elbow and the girth welding height of the elbow.

[0008] In some embodiments, determining the nozzle diameter of the elbow and the nozzle center distance of the elbow according to the bottom view image includes: rectifying distortion of the bottom view image; locating features and positions of the elbow in the bottom view image, and correcting positions of the elbow by using preset correction parameters; respectively framing an inner circle and an outer circle ROI region of a first nozzle of the elbow and an inner circle and an outer circle ROI region of a second nozzle of the elbow to determine pixel coordinates of a first nozzle center, pixel coordinates of a second nozzle center, and pixel coordinates of a nozzle diameter of the elbow; determining pixel coordinates of a nozzle center distance of the elbow according to the pixel coordinates of the first nozzle center and the pixel coordinates of the second nozzle center; and respectively converting the pixel coordinates of the nozzle diameter of the elbow and the pixel coordinates of the nozzle center distance of the elbow into world coordinates to obtain the nozzle diameter of the elbow and the nozzle center distance of the elbow.

[0009] In some embodiments, determining, according to the measurement data, whether the elbow is a qualified product or a substandard product comprises: judging the size relationship between the height of the elbow and a preset elbow height value, the size relationship between the girth weld height of the elbow and a preset girth weld height value, the size relationship between the nozzle diameter of the elbow and a preset nozzle diameter value, and the size relationship between the nozzle center distance of the elbow and a preset nozzle center distance value, respectively; if the absolute value of the difference between the height of the elbow and the preset elbow height value is less than a preset first tolerance, and the absolute value of the difference between the girth weld height of the elbow and the preset girth weld height value is less than a preset second tolerance, and the absolute value of the difference between the nozzle diameter of the elbow and the preset nozzle diameter value is less than a preset third tolerance, and the absolute value of the difference between the nozzle center distance of the elbow and the preset nozzle center distance value is less than a preset fourth tolerance, the elbow is determined to be a qualified product; if the absolute value of the difference between the height of the elbow and the preset elbow height value is greater than or equal to the preset first tolerance, or the absolute value of the difference between the girth weld height of the elbow and the preset girth weld height value is greater than or equal to the preset second tolerance, or the absolute value of the difference between the nozzle diameter of the elbow and the preset nozzle diameter value is greater than or equal to the preset third tolerance, or the absolute value of the difference between the nozzle center distance of the elbow and the preset nozzle center distance value is greater than or equal to the preset fourth tolerance, the elbow is determined to be a substandard product.

[0010] In some embodiments, the method further comprises: adjusting the preset elbow height value, the preset girth weld height value, the preset nozzle diameter value, and the preset nozzle center distance value using the measurement data after determining that the elbow is a qualified product.

[0011] Corresponding to the above method, the present application provides another aspect of an elbow screening device, which is provided with a vision system comprising: a first camera and a second camera; the first camera is used to collect a side view image of the elbow; the second camera is used to collect a top view image of the elbow; the device comprises: an acquisition unit configured to acquire the side view image of the elbow collected by the first camera and the top view image of the elbow collected by the second camera; a determination unit configured to determine measurement data of the elbow according to the side view image and the top view image; the determination unit is further configured to determine whether the elbow is a qualified product or a substandard product according to the measurement data.

[0012] In some embodiments, the measurement data comprises: the height of the elbow, the girth weld height of the elbow, the nozzle diameter of the elbow, and the nozzle center distance of the elbow; the determination unit determines the measurement data of the elbow according to the side view image and the top view image, which comprises: determining the height of the elbow and the girth weld height of the elbow according to the side view image, and determining the nozzle diameter of the elbow and the nozzle center distance of the elbow according to the top view image.

[0013] In some embodiments, the determining unit determines the height of the elbow and the girth of the elbow according to the side view image, including: rectifying distortion of the side view image; locating features and positions of the elbow in the side view image, and correcting positions of the elbow by using preset correction parameters; framing a bottom ROI region of the elbow to determine pixel coordinates of a reference straight line of the bottom of the elbow; framing a top ROI region of the elbow to determine pixel coordinates of a vertex of the elbow; respectively framing a left girth ROI region and a right girth ROI region of the elbow to determine pixel coordinates of a first point and a second point of the left girth and pixel coordinates of a first point and a second point of the right girth; inputting the pixel coordinates of the vertex of the elbow and the pixel coordinates of the reference straight line of the bottom of the elbow into a preset height calculation formula of the elbow to obtain pixel coordinates of the height of the elbow; inputting the pixel coordinates of the first point and the second point of the left girth, the pixel coordinates of the first point and the second point of the right girth, and the pixel coordinates of the reference straight line of the bottom of the elbow into a preset girth calculation formula to obtain pixel coordinates of the girth of the elbow; and respectively converting the pixel coordinates of the height of the elbow and the pixel coordinates of the girth of the elbow into world coordinates to obtain the height of the elbow and the girth of the elbow.

[0014] In some embodiments, the determining unit determines the nozzle diameter of the elbow and the nozzle center distance of the elbow according to the bottom view image, including: rectifying distortion of the bottom view image; locating features and positions of the elbow in the bottom view image, and correcting positions of the elbow by using preset correction parameters; respectively framing an inner circle ROI region and an outer circle ROI region of a first nozzle of the elbow and an inner circle ROI region and an outer circle ROI region of a second nozzle of the elbow to determine pixel coordinates of a first nozzle center, pixel coordinates of a second nozzle center, and pixel coordinates of a nozzle diameter of the elbow; determining pixel coordinates of a nozzle center distance of the elbow according to the pixel coordinates of the first nozzle center and the pixel coordinates of the second nozzle center; and respectively converting the pixel coordinates of the nozzle diameter of the elbow and the pixel coordinates of the nozzle center distance of the elbow into world coordinates to obtain the nozzle diameter of the elbow and the nozzle center distance of the elbow.

[0015] In some embodiments, the determining unit determines the elbow to be a qualified product or a substandard product according to the measurement data, including: judging the size relationship between the height of the elbow and a preset elbow height value, the size relationship between the girth weld height of the elbow and a preset girth weld height value, the size relationship between the nozzle diameter of the elbow and a preset nozzle diameter value, and the size relationship between the nozzle center distance of the elbow and a preset nozzle center distance value, respectively; if the absolute value of the difference between the height of the elbow and the preset elbow height value is less than a preset first tolerance, and the absolute value of the difference between the girth weld height of the elbow and the preset girth weld height value is less than a preset second tolerance, and the absolute value of the difference between the nozzle diameter of the elbow and the preset nozzle diameter value is less than a preset third tolerance, and the absolute value of the difference between the nozzle center distance of the elbow and the preset nozzle center distance value is less than a preset fourth tolerance, the elbow is determined to be a qualified product; if the absolute value of the difference between the height of the elbow and the preset elbow height value is greater than or equal to the preset first tolerance, or the absolute value of the difference between the girth weld height of the elbow and the preset girth weld height value is greater than or equal to the preset second tolerance, or the absolute value of the difference between the nozzle diameter of the elbow and the preset nozzle diameter value is greater than or equal to the preset third tolerance, or the absolute value of the difference between the nozzle center distance of the elbow and the preset nozzle center distance value is greater than or equal to the preset fourth tolerance, the elbow is determined to be a substandard product.

[0016] In some embodiments, the determining unit is further configured to adjust the preset elbow height value, the preset girth weld height value, the preset nozzle diameter value, and the preset nozzle center distance value using the measurement data after determining the elbow to be a qualified product.

[0017] In some embodiments, the elbow screening and classifying system is matched with the above-mentioned device.

[0018] In some embodiments, the storage medium is matched with the above-mentioned method, and the storage medium includes a stored program, wherein the program controls the device where the storage medium is located to execute the above-mentioned elbow screening method when the program is running.

[0019] In some embodiments, the computer program product is matched with the above-mentioned method, and the computer program product includes a computer program, which realizes the steps of the above-mentioned elbow screening method when the computer program product is processed and executed.

[0020] The scheme of the present application is provided with a vision system, which comprises a first camera and a second camera; the first camera is used for collecting a side view image of the elbow; the second camera is used for collecting a top view image of the elbow; measurement data of the elbow is determined according to the side view image and the top view image, and the elbow is determined as a qualified product or a defective product according to the measurement data. Thus, the actual measurement data of the elbow is calculated through the vision system, and whether the elbow is a qualified product is determined according to the actual measurement data, so that automatic classification and screening of the elbow is realized, the efficiency of automatic production of air conditioners is improved, the defective elbow is prevented from being installed on the air conditioner, and the product quality of the air conditioner is ensured to be high.

[0021] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent from the description, or can be learned by practice of the present application.

[0022] The technical scheme of the present application will be further described in detail below with the drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 A flowchart of an embodiment of the screening method of the elbow of the present application;

[0024] Figure 2 A structural schematic diagram of an embodiment of the screening device of the elbow of the present application;

[0025] Figure 3 A top view of the structure of the elbow screening and classification system of the present application;

[0026] Figure 4 A side view of the structure of the elbow screening and classification system of the present application;

[0027] Figure 5 A schematic diagram of the monitoring position of the elbow;

[0028] Figure 6 A flowchart of the elbow measurement algorithm;

[0029] Figure 7 A flowchart of the elbow screening process.

[0030] In combination with the drawings, the following are the reference signs in the embodiments of the present application:

[0031] 1-elbow; 2-inlet; 3-rotary disc; 4-classification discharge port; 5-laser sensor; 6-back side light source; 7-lower camera; 8-side camera; 9-upper light source; 102-acquisition unit; 104-determination unit. DETAILED DESCRIPTION

[0032] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described below in connection with specific embodiments of the present application and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

[0033] According to an embodiment of the present application, a screening method of an elbow is provided, which is provided with a vision system, the vision system comprising: a first camera, a second camera; the first camera is used for collecting a side view image of the elbow, such as a side camera 8; the second camera is used for collecting a top view image of the elbow, such as a bottom camera 7.

[0034] The vision system, as shown in Figure 3 and Figure 4 , comprises: a vision controller (not shown in the figure), a laser sensor 5, a side camera 8, a bottom camera 7, a rear side light source 6, an upper light source 9. The laser sensor 5 is located upstream of the incoming material, which is used to detect whether there is incoming material. Once the elbow 1 is detected, a trigger signal is generated, and the distance between the laser sensor 5 and the detected elbow 1 is detected, and the trigger signal and the distance data are sent to the vision controller. After the vision controller receives the trigger signal, it sends a camera collection signal to the side camera 8 and the bottom camera 7 at the same time. The side camera 8 collects a side view image of the elbow after receiving the collection signal, and the bottom camera 7 collects a top view image of the elbow after receiving the collection signal. The rear side light source 6 and the upper light source 9 are used to provide illumination when collecting the side view image and the top view image. Then the side view image and the top view image are sent to the vision controller to calculate the measurement data of the elbow in the image. The distance detected by the laser sensor 5 is the distance from the plane where the laser sensor 5 and the side camera 8 are located to the plane where the elbow is located.

[0035] After the elbow is placed on the counterclockwise rotating rotary disc 3 through the feeding port 1, as the rotary disc 3 rotates, the elbow is detected by the laser sensor 5 when it rotates to the front of the laser sensor 5, triggering the vision system and calculating the measurement data of the elbow through the collected image. The measurement data is used to determine whether the elbow is a qualified product or a defective product. When the elbow rotates to the classification discharge port 4 along with the rotary disc, the airflow intensity and direction of the blower organ are controlled to discharge the qualified products and the defective products from different discharge ports, realizing automatic data measurement, screening and classification of the elbow.

[0036] As shown in Figure 1 , the flowchart of an embodiment of the method of the present application. The screening method of the elbow can comprise: steps S110 to S130.

[0037] In step S110, the side view image of the bend captured by the first camera and the bottom view image of the bend captured by the second camera are obtained.

[0038] In some embodiments, the measurement data includes: the height of the elbow, the circumferential weld height of the elbow, the pipe diameter of the elbow, and the center distance between the pipe openings of the elbow.

[0039] like Figure 5 The side and top views of the elbow are shown. The height of the elbow is L, the height of the circumferential weld is H, the diameter of the pipe opening is d, and the distance between the center of the pipe opening is P. Additionally, p0 is the apex of the elbow, p1 is the lower left edge of the left circumferential weld, p2 is the lower right edge of the left circumferential weld, p3 is the lower left edge of the right circumferential weld, p4 is the lower right edge of the right circumferential weld, p5 is the center of the left circumferential weld, p6 is the center of the right circumferential weld, and l1 is the reference straight line at the bottom of the elbow.

[0040] In step S120, the measurement data of the elbow are determined based on the side view image and the bottom view image.

[0041] In some embodiments, the specific process of determining the measurement data of the elbow based on the side view image and the bottom view image in step S120 includes: determining the height of the elbow and the circumferential weld height of the elbow based on the side view image, and determining the pipe diameter of the elbow and the center distance of the pipe opening of the elbow based on the bottom view image.

[0042] Before determining the measurement data of the elbow based on the image, it is necessary to calculate the preset correction parameters. Since the measurement data of the elbow obtained by the measurement algorithm in the vision system is in pixel coordinates, it needs to be converted into the true value in world coordinates. This conversion mapping relationship is completed through camera calibration and coordinate transformation. Since there is a certain distance deviation between the elbow plane and the calibration plane during the elbow receiving process, it is necessary to correct this distance deviation when determining the elbow measurement data based on the image. Therefore, the preset correction parameters are used to correct this distance deviation. That is, the function of the preset correction parameters is to correct the distance deviation between the elbow plane and the calibration plane, so that the measurement data is closer to the physical true value of the elbow.

[0043] The process of calculating the preset correction parameters is as follows: based on the side view and top view images, the distance data of the bend obtained by the laser sensor is fused, the measurement error is minimized using the least squares method, and the feature point positions obtained from image processing are used. and laser sensor Find the optimal fusion value based on the measured distance. The optimal fusion value is the preset correction parameter. The formula for minimizing the sum of squared errors is:

[0044]

[0045] The fusion value of each measurement point is calculated according to the optimal solution a and b obtained by the minimum error square sum, and the formula is:

[0046]

[0047] The optimal fusion value is fed back to the vision system, and the height of the elbow, the ring welding height, the pipe mouth diameter, the pipe mouth center distance and other data are measured.

[0048] In some embodiments, the specific process of determining the height of the elbow and the ring welding height of the elbow according to the side view image includes steps S210 to S280.

[0049] Step S210, distortion correction is performed on the side view image.

[0050] Due to the vibration of the rotating disc guide rail, the plane where the elbow is located and the camera calibration plane may not be in the same plane, and there is a certain distance error. Therefore, it is necessary to fuse the distance data collected by the laser sensor to proportionally enlarge and crop the image. At the same time, since the position of the elbow in the camera field of view is not fixed, and the camera has a certain degree of distortion, it is necessary to further correct the image coordinates. The distortion correction formula is as follows:

[0051]

[0052] Where (x, y) represents the image coordinates before distortion, (x d ,y d ) represents the image coordinates after distortion, k1, k2, k3, p1, p2 represent the distortion coefficients.

[0053] Step S220, in the side view image, the features and positions of the elbow are located, and the position of the elbow is corrected using the preset correction parameters.

[0054] The template image of the small elbow position is framed, the feature is roughly positioned, the approximate position of the elbow in the image is determined, and the template matching score match_score is output. At the same time, the position of the elbow is corrected using the preset correction parameters calculated before.

[0055] Step S230, the bottom ROI region of the elbow is framed, and the pixel coordinates of the reference straight line of the bottom of the elbow are determined.

[0056] ​The Region of Interest (ROI) at the bottom of the small elbow pipe opening is selected. A pixel-level edge detection algorithm is used to extract pixel-level edge information for coarse edge localization. Then, a surface fitting method is used to further subdivide the pixel-level edge information to obtain sub-pixel edge coordinates for fine edge localization. Finally, iterative reweighted least squares line fitting is used to iteratively fit the sub-pixel edge coordinates, ultimately completing the detection of the baseline straight line at the bottom of the elbow. The baseline straight line is shown below. Figure 5 l1 in the middle.

[0057] Step S240: Select the top ROI region of the bend and determine the pixel coordinates of the bend's vertex.

[0058] The apex of the bend is as follows Figure 5 Point p0 in the diagram. Select the ROI area at the top of the bend. Within this area, place caliper tools at equal intervals with the ROI angle. Use the caliper tools to detect sub-pixel edge points within the area and calculate the distance of each edge point relative to the ROI starting point. The closest point is the bend vertex.

[0059] Step S250: Select the left and right circumferential weld ROI areas of the elbow respectively, and determine the pixel coordinates of the first and second points of the left circumferential weld, and the pixel coordinates of the first and second points of the right circumferential weld.

[0060] like Figure 5 As shown, the first and second points of the left circumferential weld are p1 and p2, respectively, and the first and second points of the right circumferential weld are p3 and p4, respectively. The ROI region of the left and right circumferential welds of the elbow is selected. Within this region, an edge detection algorithm is used to detect the lower edge points of the left and right circumferential welds, and the distance of each edge point relative to the bottom reference line is calculated. The point with the closest distance is the lower edge point of the small elbow weld.

[0061] Step S260: Substitute the pixel coordinates of the vertex of the bend and the pixel coordinates of the reference line at the bottom of the bend into the preset bend height calculation formula to obtain the pixel coordinates of the bend height.

[0062] The formula for calculating the elbow height is:

[0063] L=DistancePointLine(p0,l0);

[0064] L represents the pixel coordinates of the elbow's height, p0 represents the pixel coordinates of the elbow's apex, and l0 represents the pixel coordinates of the baseline at the bottom of the elbow.

[0065] Step S270, the pixel coordinates of the first point and the second point of the left circumferential weld, the pixel coordinates of the first point and the second point of the right circumferential weld, the pixel coordinates of the reference straight line of the elbow bottom are brought into a preset circumferential weld height calculation formula to obtain the pixel coordinates of the circumferential weld height of the elbow. The circumferential weld height calculation formula is:

[0066]

[0067] H is the pixel coordinates of the circumferential weld height of the elbow, p1 is the pixel coordinates of the first point of the left circumferential weld, p2 is the pixel coordinates of the second point of the left circumferential weld, p3 is the pixel coordinates of the first point of the right circumferential weld, p4 is the pixel coordinates of the second point of the right circumferential weld, and l0 is the pixel coordinates of the reference straight line of the elbow bottom.

[0068] Step S280, the pixel coordinates of the height of the elbow and the pixel coordinates of the circumferential weld height of the elbow are respectively converted into world coordinates to obtain the height of the elbow and the circumferential weld height of the elbow.

[0069] After obtaining the pixel coordinates of the measurement data, the pixel coordinates are converted into world coordinates and output, and then it is judged whether the elbow is a qualified product.

[0070] In some embodiments, the specific process of determining the pipe mouth diameter of the elbow and the pipe mouth center distance of the elbow according to the downward image includes steps S310 to S350.

[0071] Step S310, distortion correction is performed on the downward image.

[0072] The process of performing distortion correction on the downward image is the same as that of performing distortion correction on the side image, which will not be repeated here.

[0073] Step S320, the features and positions of the elbow are located in the downward image, and the positions of the elbow are corrected by using a preset correction parameter.

[0074] The way of locating the features and positions of the elbow in the downward image is the same as that of locating the features and positions of the elbow in the side image, which will not be repeated here.

[0075] Step S330, the inner circle and outer circle ROI regions of the first pipe mouth and the inner circle and outer circle ROI regions of the second pipe mouth of the elbow are respectively framed, the pixel coordinates of the first pipe mouth center, the pixel coordinates of the second pipe mouth center, and the pixel coordinates of the pipe mouth diameter of the elbow are determined.

[0076] Step S340, the pixel coordinates of the pipe mouth center distance of the elbow are determined according to the pixel coordinates of the first pipe mouth center and the pixel coordinates of the second pipe mouth center.

[0077] As shown in Figure 5 , the first pipe orifice center is p5, the second pipe orifice center is p6, and the pipe orifice diameter of the elbow is d. The inner circle and the outer circle ROI region of the small elbow orifice are framed, and the standard circle fitting is performed according to the circular edge point coordinates in the framed annular ROI region. The algorithm principle is the same as that of the elbow bottom reference straight line detection, and only the least square fitting method is different. The circle equation is as follows:

[0078] (x-a) 2 +(y-b) 2 =r 2 ;

[0079] The least square circle fitting calculation formula is:

[0080]

[0081] Wherein, r is the radius of the pipe orifice, and the pipe orifice diameter d = 2r. The calculation formula of the center distance P is:

[0082] P = DistancePointPoint (p5, p6);

[0083] P5 is the pixel coordinate of the first pipe orifice center, and p6 is the pixel coordinate of the second pipe orifice center.

[0084] Step S350, respectively, the pixel coordinates of the pipe orifice diameter of the elbow, the pixel coordinates of the pipe orifice center distance of the elbow are converted into world coordinates, and the pipe orifice diameter of the elbow and the pipe orifice center distance of the elbow are obtained.

[0085] The rotation matrix, translation vector, camera intrinsic matrix, distortion coefficient and other data calculated by the camera calibration algorithm are used to convert the measurement data in pixel coordinates measured by the vision measurement algorithm into real physical world values.

[0086] By fusing laser sensor data and image data, using intelligent sensing technology and high resolution image processing technology, the real value of the elbow size on the flow line is realized.

[0087] Figure 6 The flow chart of the elbow measurement algorithm is shown in Figure 6 , which comprises:

[0088] Step 1, after the glass turntable is fed, when the laser sensor detects the elbow, the side camera and the lower camera are triggered to take pictures, and the side view image and the overhead view image are obtained.

[0089] Step 2, for the side view image, the side view image is rectified for distortion; a ROI region of the elbow is framed, template matching is performed, coarse positioning of features is performed, the approximate position of the elbow in the image is determined, and position correction is performed; a ROI region of the bottom of the pipe opening of the elbow is framed, a reference straight line at the bottom of the elbow is detected; a ROI region of the top of the elbow is framed, the top of the elbow is detected, and the height L of the elbow is calculated; ROI regions of the left and right welding rings are framed, the lower edge of the welding ring is detected, and the height H of the welding ring is calculated; the pixel coordinates of the height L of the elbow and the height H of the welding ring are converted into world coordinates, and the world coordinates of the height L of the elbow and the height H of the welding ring are output.

[0090] Step 3, for the top view image, the top view image is rectified for distortion; a ROI region of the elbow is framed, template matching is performed, coarse positioning of features is performed, the approximate position of the elbow in the image is determined, and position correction is performed; a ROI region of the inside and outside of the pipe opening is framed, the pipe opening circle is detected, the left and right pipe opening diameters d are calculated, the left and right pipe opening center coordinates are determined, and the pipe opening center distance P is calculated; the pixel coordinates of the left and right pipe opening diameters d and the pipe opening center distance P are converted into world coordinates, and the world coordinates of the left and right pipe opening diameters d and the pipe opening center distance P are output.

[0091] At step S130, it is determined according to the measurement data whether the elbow is a qualified product or a defective product.

[0092] The scheme calculates the actual measurement data of the elbow through the visual system, judges whether the elbow is a qualified product according to the actual measurement data, realizes automatic classification and screening of the elbow, does not need to manually switch projects, reduces manual intervention, improves the efficiency of automatic production of air conditioners, prevents defective elbows from being installed on air conditioners, and ensures that air conditioners have high product quality.

[0093] In some embodiments, in step S130, the specific process of determining whether the elbow is a qualified product or a defective product according to the measurement data includes: respectively judging the size relationship between the height of the elbow and the preset elbow height value, the size relationship between the girth welding height of the elbow and the preset girth welding height value, the size relationship between the pipe opening diameter of the elbow and the preset pipe opening diameter value, and the size relationship between the pipe opening center distance of the elbow and the preset pipe opening center distance value; if the absolute value of the difference between the height of the elbow and the preset elbow height value is less than a preset first tolerance, and the absolute value of the difference between the girth welding height of the elbow and the preset girth welding height value is less than a preset second tolerance, and the absolute value of the difference between the pipe opening diameter of the elbow and the preset pipe opening diameter value is less than a preset third tolerance, and the absolute value of the difference between the pipe opening center distance of the elbow and the preset pipe opening center distance value is less than a preset fourth tolerance, the elbow is determined to be a qualified product; if the absolute value of the difference between the height of the elbow and the preset elbow height value is greater than or equal to the preset first tolerance, or the absolute value of the difference between the girth welding height of the elbow and the preset girth welding height value is greater than or equal to the preset second tolerance, or the absolute value of the difference between the pipe opening diameter of the elbow and the preset pipe opening diameter value is greater than or equal to the preset third tolerance, or the absolute value of the difference between the pipe opening center distance of the elbow and the preset pipe opening center distance value is greater than or equal to the preset fourth tolerance, the elbow is determined to be a defective product.

[0094] The current small elbow is respectively matched with five standard template elbows, and the elbow model is distinguished according to the score, and the higher the score, the more matched the elbow model. Specifically, the current elbow measurement value is compared with the preset elbow standard value to judge whether the elbow height is less than the tolerance value, whether the left and right welding rings of the elbow are missing, whether the left and right outer welding ring height is less than the tolerance, whether the pipe opening diameter is less than the tolerance, and whether the pipe opening center distance is less than the tolerance, and if all the requirements are met, the current elbow is a qualified product, and its measurement value is recorded; otherwise, the current elbow is a defective product, and the material is prepared to be discarded.

[0095] In some embodiments, further comprising: after determining that the elbow is a qualified product, adjusting the preset elbow height value, the preset girth welding height value, the preset pipe opening diameter value, and the preset pipe opening center distance value using the measurement data.

[0096] Due to the gap in the manufacturing process, the real value distribution of each batch of elbow incoming material has a certain gap with the standard value of the drawing, therefore, the average value and the standard deviation of the cumulative measurement parameters (including the elbow height, the welding ring height, the pipe opening diameter, and the pipe opening center distance) of the current elbow model are calculated, the Shapiro-Wilk test is used to test whether the cumulative measurement data conforms to the normal distribution, and the results are fed back in real time and the screening parameter standard value and the qualified range of each batch of elbow are dynamically adjusted according to the results.

[0097] By adaptively adjusting the screening parameter standard and the qualified range according to the real-time measurement results, the differences of different batches of materials are adapted, the small elbows are reasonably separated according to the size and shape of different models, and the small elbow materials are more efficiently and intelligently screened and classified.

[0098] Figure 7 A flowchart of the elbow screening process is shown in FIG. 1, and the method comprises the following steps. Figure 7

[0099] Step 11: input the measurement value of the elbow, and initialize the standard value of the elbow.

[0100] Step 12: when the welding ring does not exist missing, and the absolute value of the difference between the elbow height, the elbow welding ring, the left and right welding ring height, the pipe opening diameter, and the pipe opening center distance and the standard value is less than the tolerance, the elbow is a qualified product, and the average value and the standard deviation of the cumulative measurement parameters of the current model elbow are calculated according to the measurement value of the elbow, and the standard value of the elbow is dynamically adjusted.

[0101] Step 13: when the welding ring exists missing, or the absolute value of the difference between any one of the elbow height, the elbow welding ring, the left and right welding ring height, the pipe opening diameter, and the pipe opening center distance and the standard value is greater than the tolerance, the elbow is a defective product, and the elbow is discarded.

[0102] The technical scheme of the embodiment is provided with a vision system, which includes a first camera and a second camera; the first camera is used to collect a side view image of the elbow; the second camera is used to collect a top view image of the elbow; the measurement data of the elbow is determined according to the side view image and the top view image, and the elbow is determined to be a qualified product or a defective product according to the measurement data. Thus, the actual measurement data of the elbow is calculated through the vision system, and whether the elbow is a qualified product is determined according to the actual measurement data, so that automatic classification and screening of the elbow are realized, the efficiency of automatic production of air conditioners is improved, the defective elbow is prevented from being installed on the air conditioner, and the product quality of the air conditioner is ensured to be high.

[0103] According to the embodiment of the present application, a screening device for an elbow corresponding to the screening method of the elbow is also provided. The vision system is provided with a first camera and a second camera; the first camera is used to collect a side view image of the elbow, such as a side camera 8; the second camera is used to collect a top view image of the elbow, such as a lower camera 7.

[0104] The vision system is as shown in FIG. 2. Figure 3 and Figure 4 ​As shown, it comprises: a visual controller (not shown in the figure), a laser sensor 5, a side camera 8, a lower camera 7, a rear light source 6, an upper light source 9. The laser sensor 5 is located upstream of the incoming material, used to detect whether there is incoming material, once the elbow 1 is detected, a trigger signal is generated, and the distance between the laser sensor 5 and the detected elbow 1 is detected, and the trigger signal and distance data are sent to the visual controller. After the visual controller receives the trigger signal, it sends a camera acquisition signal to the side camera 8 and the lower camera 7 at the same time, the side camera 8 receives the acquisition signal and acquires the side view image of the elbow, and the lower camera 7 receives the acquisition signal and acquires the overhead view image of the elbow. The rear light source 6 and the upper light source 9 are used to provide illumination when acquiring the side view image and the overhead view image. Then the side view image and the overhead view image are sent to the visual controller to calculate the measurement data of the elbow in the image. Wherein the distance detected by the laser sensor 5 is the distance from the plane where the laser sensor 5 and the side camera 8 are located to the plane where the elbow is located.

[0105] After the elbow is placed on the rotating disc 3 rotating counterclockwise through the feed port 1, as the rotating disc 3 rotates, when the elbow rotates to the front of the laser sensor 5, it is detected by the laser sensor 5, triggering the vision system and calculating the measurement data of the elbow through the acquired image, judging whether the elbow is a qualified product or a defective product through the measurement data, when the elbow rotates to the classification discharge port 4 with the rotating disc, the airflow intensity and direction of the blower organ are controlled to discharge the qualified products and defective products from different discharge ports, realizing automatic data measurement, screening and classification of the elbow.

[0106] Referring to Figure 2 The structure schematic diagram of an embodiment of the device of the application is shown. The elbow screening device can comprise: an acquisition unit 102 and a determination unit 104.

[0107] The acquisition unit 102 is configured to acquire the side view image of the elbow acquired by the first camera and the overhead view image of the elbow acquired by the second camera. The specific functions and processes of the acquisition unit 102 are described in step S110.

[0108] In some embodiments, the measurement data comprises: the height of the elbow, the girth welding height of the elbow, the pipe opening diameter of the elbow, and the pipe opening center distance of the elbow.

[0109] As Figure 5 The side view and overhead view of the elbow are shown, the height of the elbow is L, the girth welding height of the elbow is H, the pipe opening diameter of the elbow is d, and the pipe opening center distance of the elbow is P. In addition, p0 is the vertex of the elbow, p1 is the left girth left lower edge point, p2 is the left girth right lower edge point, p3 is the right girth left lower edge point, p4 is the right girth right lower edge point, p5 is the left girth center, p6 is the right girth center, and l1 is the elbow bottom reference straight line.

[0110] The determining unit 104 is configured to determine the measurement data of the bend based on the side view image and the bottom view image. The specific functions and processing of the determining unit 104 are described in step S120.

[0111] In some embodiments, the determining unit 104 determines the specific process of measuring the elbow based on the side view image and the bottom view image, including: determining the height of the elbow and the circumferential weld height of the elbow based on the side view image, and determining the pipe diameter of the elbow and the distance between the pipe openings of the elbow based on the bottom view image.

[0112] Before determining the measurement data of the elbow based on the image, it is necessary to calculate the preset correction parameters. Since the measurement data of the elbow obtained by the measurement algorithm in the vision system is in pixel coordinates, it needs to be converted into the true value in world coordinates. This conversion mapping relationship is completed through camera calibration and coordinate transformation. Since there is a certain distance deviation between the elbow plane and the calibration plane during the elbow receiving process, it is necessary to correct this distance deviation when determining the elbow measurement data based on the image. Therefore, the preset correction parameters are used to correct this distance deviation. That is, the function of the preset correction parameters is to correct the distance deviation between the elbow plane and the calibration plane, so that the measurement data is closer to the physical true value of the elbow.

[0113] The process of calculating the preset correction parameters is as follows: based on the side view and top view images, the distance data of the bend obtained by the laser sensor is fused, the measurement error is minimized using the least squares method, and the feature point positions obtained from image processing are used. and laser sensor Find the optimal fusion value based on the measured distance. The optimal fusion value is the preset correction parameter. The formula for minimizing the sum of squared errors is:

[0114]

[0115] The fusion value for each measurement point is calculated based on the optimal solutions a and b obtained from the minimum sum of squared errors, using the following formula:

[0116]

[0117] Then the optimal fusion value The data is fed back to the vision system to measure the height of the elbow, the height of the circumferential weld, the diameter of the pipe opening, and the distance between the center of the pipe opening.

[0118] In some embodiments, the specific process by which the determining unit 104 determines the height of the elbow and the circumferential weld height of the elbow based on the side view image includes:

[0119] The determining unit 104 is further configured to correct distortion of the side-view image.

[0120] Due to the vibration of the rotating disc guide rail, the plane where the elbow is located and the plane where the camera is calibrated may not be in the same plane, and there is a certain distance error. Therefore, it is necessary to fuse the distance data collected by the laser sensor to proportionally enlarge and crop the image. At the same time, since the position of the elbow in the field of view of the camera is not fixed, and the camera has a certain degree of distortion, it is necessary to further correct the image coordinates. The distortion correction formula is as follows:

[0121]

[0122] where (x, y) represents the image coordinates before distortion, (x d ,y d ) represents the image coordinates after distortion, k1, k2, k3, p1, and p2 represent the distortion coefficients.

[0123] The determining unit 104 is further configured to locate the features and position of the elbow in the side-view image, and correct the position of the elbow by using a preset correction parameter.

[0124] The template image where the small elbow is located is framed, the feature is roughly positioned, the approximate position of the elbow in the image is determined, and a template matching score match_score is output. At the same time, the position of the elbow is corrected by using the preset correction parameter calculated before.

[0125] The determining unit 104 is further configured to frame the ROI region of the bottom of the elbow, and determine the pixel coordinates of the reference straight line of the bottom of the elbow.

[0126] The ROI region of the bottom of the pipe opening of the small elbow is framed, the pixel-level edge information of the image is extracted by a pixel-level edge detection algorithm to complete rough positioning of the edge points, the pixel-level edge information is subdivided according to a curved surface fitting method to obtain sub-pixel edge coordinates to complete precise positioning of the edge, and finally the sub-pixel edge point coordinates are iteratively fitted by an iteratively reweighted least squares straight line fitting, so as to finally complete detection of the reference straight line of the bottom of the elbow, which is l1 in Figure 5 .

[0127] The determining unit 104 is further configured to frame the ROI region of the top of the elbow, and determine the pixel coordinates of the vertex of the elbow.

[0128] The vertex of the elbow is as shown in Figure 5Point p0 in the diagram. Select the ROI area at the top of the bend. Within this area, place caliper tools at equal intervals with the ROI angle. Use the caliper tools to detect sub-pixel edge points within the area and calculate the distance of each edge point relative to the ROI starting point. The closest point is the bend vertex.

[0129] The determining unit 104 is further configured to select the left circumferential weld ROI region and the right circumferential weld ROI region of the elbow respectively, and determine the pixel coordinates of the first point and the second point of the left circumferential weld, and the pixel coordinates of the first point and the second point of the right circumferential weld.

[0130] like Figure 5 As shown, the first and second points of the left circumferential weld are p1 and p2, respectively, and the first and second points of the right circumferential weld are p3 and p4, respectively. The ROI region of the left and right circumferential welds of the elbow is selected. Within this region, an edge detection algorithm is used to detect the lower edge points of the left and right circumferential welds, and the distance of each edge point relative to the bottom reference line is calculated. The point with the closest distance is the lower edge point of the small elbow weld.

[0131] The determining unit 104 is further configured to input the pixel coordinates of the vertex of the bend and the pixel coordinates of the reference line at the bottom of the bend into a preset bend height calculation formula to obtain the pixel coordinates of the bend height.

[0132] The formula for calculating the elbow height is:

[0133] L=DistancePointLine(p0,l0);

[0134] L represents the pixel coordinates of the elbow's height, p0 represents the pixel coordinates of the elbow's apex, and l0 represents the pixel coordinates of the baseline at the bottom of the elbow.

[0135] The determining unit 104 is further configured to substitute the pixel coordinates of the first and second points of the left circumferential weld, the pixel coordinates of the first and second points of the right circumferential weld, and the pixel coordinates of the reference line at the bottom of the elbow into a preset circumferential weld height calculation formula to obtain the pixel coordinates of the circumferential weld height of the elbow. The circumferential weld height calculation formula is:

[0136]

[0137] H represents the pixel coordinates of the circumferential weld height of the elbow, p1 represents the pixel coordinates of the first point of the left circumferential weld, p2 represents the pixel coordinates of the second point of the left circumferential weld, p3 represents the pixel coordinates of the first point of the right circumferential weld, p4 represents the pixel coordinates of the second point of the right circumferential weld, and l0 represents the pixel coordinates of the reference straight line at the bottom of the elbow.

[0138] The determining unit 104 is further configured to respectively convert the pixel coordinates of the height of the elbow and the pixel coordinates of the girth of the elbow into world coordinates to obtain the height of the elbow and the girth of the elbow.

[0139] After obtaining the pixel coordinates of the measurement data, the pixel coordinates are converted into world coordinates and output, and then it is determined whether the elbow is a qualified product.

[0140] In some embodiments, the determining unit 104 determines the pipe mouth diameter of the elbow and the pipe mouth center distance of the elbow according to the top view image, and the specific process includes:

[0141] The determining unit 104 is further configured to perform distortion correction on the top view image.

[0142] The process of performing distortion correction on the top view image is the same as that of performing distortion correction on the side view image, which will not be repeated here.

[0143] The determining unit 104 is further configured to locate the features and positions of the elbow in the top view image, and correct the positions of the elbow by using a preset correction parameter.

[0144] The way of locating the features and positions of the elbow in the top view image is the same as that in the side view image, which will not be repeated here.

[0145] The determining unit 104 is further configured to respectively frame the inner circle and outer circle ROI regions of the first pipe mouth and the inner circle and outer circle ROI regions of the second pipe mouth of the elbow, and determine the pixel coordinates of the center of the first pipe mouth, the pixel coordinates of the center of the second pipe mouth, and the pixel coordinates of the pipe mouth diameter of the elbow.

[0146] The determining unit 104 is further configured to determine the pixel coordinates of the pipe mouth center distance of the elbow according to the pixel coordinates of the center of the first pipe mouth and the pixel coordinates of the center of the second pipe mouth.

[0147] As shown in Figure 5 The center of the first pipe mouth is p5, the center of the second pipe mouth is p6, and the pipe mouth diameter of the elbow is d. The inner circle and outer circle ROI regions of the small elbow pipe mouth are framed, and standard circle fitting is performed according to the circle edge point coordinates in the framed annular ROI region. The algorithm principle is the same as that of the elbow bottom reference straight line detection, and only the least square fitting method is different. The circle equation is as follows:

[0148] (x-a) 2 +(y-b) 2 =r 2 ;

[0149] The least square circle fitting calculation formula is:

[0150]

[0151] wherein r is the radius of the pipe opening, the pipe opening diameter d = 2r. The formula for calculating the center distance P is:

[0152] P = DistancePointPoint(p5, p6);

[0153] p5 is the pixel coordinate of the first pipe opening center, and p6 is the pixel coordinate of the second pipe opening center.

[0154] The determination unit 104 is specifically further configured to respectively convert the pixel coordinates of the pipe opening diameter of the elbow and the pixel coordinates of the pipe opening center distance of the elbow into world coordinates to obtain the pipe opening diameter of the elbow and the pipe opening center distance of the elbow.

[0155] The rotation matrix, the translation vector, the camera intrinsic matrix, the distortion coefficient and the like calculated by using the camera calibration algorithm are used to convert the measurement data in the pixel coordinates measured by the vision measurement algorithm into real physical world values.

[0156] By fusing the laser sensor data and the image data, the method combining the intelligent sensing technology and the high-resolution image processing technology is used to realize the real-time online measurement of the real value of the elbow size on the flow line.

[0157] Figure 6 The flowchart of the elbow measurement algorithm is shown in FIG. 1, and the method comprises the following steps. Figure 6

[0158] Step 1: After the glass turntable is loaded, when the laser sensor detects the elbow, the side camera and the lower camera are triggered to take pictures to obtain the side view image and the upward view image.

[0159] Step 2: For the side view image, the side view image is corrected for distortion; the ROI region of the elbow is framed and selected, template matching is performed, coarse positioning of the features is performed, the approximate position of the elbow in the image is determined, and position correction is performed; the ROI region of the bottom of the pipe opening of the elbow is framed and selected, the reference straight line at the bottom of the elbow is detected; the ROI region of the top of the elbow is framed and selected, the top of the elbow is detected, and the height L of the elbow is calculated; the ROI regions of the left and right welding rings are framed and selected, the lower edge of the welding ring is detected, and the height H of the welding ring is calculated; the pixel coordinates of the height L of the elbow and the height H of the welding ring are converted into world coordinates, and the world coordinates of the height L of the elbow and the height H of the welding ring are output.

[0160] ​Step 3, for the top view image, the top view image is rectified; the elbow ROI region is framed and selected, template matching is performed, coarse positioning of features is performed, the approximate position of the elbow in the image is determined, and position correction is performed; the ROI region of the pipe opening inside and outside the original is framed and selected, the pipe opening circle is detected, the left and right pipe opening diameters d are calculated, the left and right pipe opening center coordinates are determined, and the pipe opening center distance P is calculated; the pixel coordinates of the left and right pipe opening diameters d and the pipe opening center distance P are converted into world coordinates, and the world coordinates of the left and right pipe opening diameters d and the pipe opening center distance P are output.

[0161] The determination unit 104 is further configured to determine, according to the measurement data, whether the elbow is a qualified product or a substandard product. For specific functions and processes of the determination unit 104, refer to step S130.

[0162] The actual measurement data of the elbow is calculated through the visual system, and whether the elbow is a qualified product is determined according to the actual measurement data, so that automatic classification and screening of the elbow are realized, manual switching of engineering is not needed, manual intervention is reduced, the efficiency of automatic production of air conditioners is improved, the elbow of a substandard product is prevented from being installed on an air conditioner, and the product quality of the air conditioner is ensured to be high.

[0163] In some embodiments, the specific process of the determination unit 104 for determining, according to the measurement data, whether the elbow is a qualified product or a substandard product includes: respectively judging the size relationship between the height of the elbow and a preset elbow height value, the size relationship between the girth welding height of the elbow and a preset girth welding height value, the size relationship between the pipe opening diameter of the elbow and a preset pipe opening diameter value, and the size relationship between the pipe opening center distance of the elbow and a preset pipe opening center distance value; if the absolute value of the difference between the height of the elbow and the preset elbow height value is less than a preset first tolerance, and the absolute value of the difference between the girth welding height of the elbow and the preset girth welding height value is less than a preset second tolerance, and the absolute value of the difference between the pipe opening diameter of the elbow and the preset pipe opening diameter value is less than a preset third tolerance, and the absolute value of the difference between the pipe opening center distance of the elbow and the preset pipe opening center distance value is less than a preset fourth tolerance, the elbow is determined to be a qualified product; if the absolute value of the difference between the height of the elbow and the preset elbow height value is greater than or equal to the preset first tolerance, or the absolute value of the difference between the girth welding height of the elbow and the preset girth welding height value is greater than or equal to the preset second tolerance, or the absolute value of the difference between the pipe opening diameter of the elbow and the preset pipe opening diameter value is greater than or equal to the preset third tolerance, or the absolute value of the difference between the pipe opening center distance of the elbow and the preset pipe opening center distance value is greater than or equal to the preset fourth tolerance, the elbow is determined to be a substandard product.

[0164] The current small elbow is matched with five standard template elbows respectively, and the elbow model is distinguished according to the score, and the higher the score is, the more matched the elbow model is. Specifically, the current elbow measurement value is compared with the preset elbow standard value, whether the elbow height is less than the tolerance value, whether the left and right welding rings exist, whether the left and right outer welding ring height is less than the tolerance, whether the pipe diameter is less than the tolerance, and whether the pipe center distance is less than the tolerance, if all meet the requirements, the current elbow is a qualified product, and the measurement value is recorded; otherwise, the current elbow is a defective product, and the material is prepared to be discarded.

[0165] In some embodiments, the determination unit 104 is also configured to adjust the preset elbow height value, the preset ring welding height value, the preset pipe diameter value, and the preset pipe center distance value using the measurement data after determining that the elbow is a qualified product.

[0166] Due to the difference in manufacturing process, the real value distribution of each batch of elbow material has a certain gap with the standard value of the drawing, therefore, the average value and the standard deviation of the cumulative measurement parameters (including the elbow height, the welding ring height, the pipe diameter, and the pipe center distance) of the model to which the current elbow belongs are calculated, the Shapiro-Wilk test is used to test whether the cumulative measurement data conforms to the normal distribution, and the screening parameter standard value and the qualified range of each batch of elbow are dynamically adjusted according to the results.

[0167] By adaptively adjusting the screening parameter standard and the qualified range according to the real-time measurement results, the differences of different batches of materials are adapted, and the small elbows are reasonably separated according to the size and shape of different models to separate the qualified products and the defective products, and the small elbows are more efficiently and intelligently screened and classified.

[0168] Figure 7 The flowchart of the elbow screening process is shown in FIG. 1, and the method comprises the following steps. Figure 7

[0169] Step 11, input the measurement value of the elbow, and initialize the standard value of the elbow.

[0170] Step 12, when the welding ring does not exist, and the absolute value of the difference between the elbow height, the elbow welding ring, the left and right welding ring height, the pipe diameter, and the pipe center distance and the standard value is less than the tolerance, the elbow is a qualified product, and the average value and the standard deviation of the cumulative measurement parameters of the current model elbow are calculated according to the measurement value of the elbow, and the standard value of the elbow is dynamically adjusted.

[0171] Step 13, when the welding ring exists, or the absolute value of the difference between any one of the elbow height, the elbow welding ring, the left and right welding ring height, the pipe diameter, and the pipe center distance and the standard value is greater than the tolerance, the elbow is a defective product, and the elbow is discarded.

[0172] ​Since the processing and functions realized by the device of the embodiment are basically corresponding to the embodiments, principles and examples of the foregoing method, the description of the embodiment does not elaborate on the related descriptions in the foregoing embodiments, which are not repeated here.

[0173] According to the technical solution of the present application, the visual system is provided, and the visual system comprises: a first camera and a second camera; the first camera is used for collecting a side view image of the elbow; the second camera is used for collecting a top view image of the elbow; measurement data of the elbow is determined according to the side view image and the top view image, and the elbow is determined as a qualified product or a defective product according to the measurement data. Thus, the actual measurement data of the elbow is calculated through the visual system, whether the elbow is a qualified product is determined according to the actual measurement data, automatic classification and screening of the elbow are realized, the efficiency of automatic production of the air conditioner is improved, the defective elbow is prevented from being installed on the air conditioner, and the product quality of the air conditioner is ensured to be high.

[0174] According to the embodiment of the present application, a kind of elbow screening classification system corresponding to the screening device of elbow is also provided.The elbow screening classification system can include: the screening device of elbow described above.

[0175] Since the processing and functions realized by the elbow screening classification system of the embodiment are basically corresponding to the embodiments, principles and examples of the foregoing device, the description of the embodiment does not elaborate on the related descriptions in the foregoing embodiments, which are not repeated here.

[0176] According to the technical solution of the present application, the visual system is provided, and the visual system comprises: a first camera and a second camera; the first camera is used for collecting a side view image of the elbow; the second camera is used for collecting a top view image of the elbow; measurement data of the elbow is determined according to the side view image and the top view image, and the elbow is determined as a qualified product or a defective product according to the measurement data. Thus, the actual measurement data of the elbow is calculated through the visual system, whether the elbow is a qualified product is determined according to the actual measurement data, automatic classification and screening of the elbow are realized, the efficiency of automatic production of the air conditioner is improved, the defective elbow is prevented from being installed on the air conditioner, and the product quality of the air conditioner is ensured to be high.

[0177] According to the embodiment of the present application, a kind of storage medium corresponding to the screening method of elbow is also provided, and the storage medium includes stored program, wherein, when the program runs, the device where the storage medium is located is controlled to execute the screening method of elbow described above.

[0178] Since the processing and functions realized by the storage medium of the embodiment are basically corresponding to the embodiments, principles and examples of the foregoing method, the description of the embodiment does not elaborate on the related descriptions in the foregoing embodiments, which are not repeated here.

[0179] The technical scheme of the present application is provided with a vision system, which comprises a first camera and a second camera; the first camera is used for collecting a side view image of the elbow; the second camera is used for collecting a top view image of the elbow; measurement data of the elbow is determined according to the side view image and the top view image, and the elbow is determined as a qualified product or a defective product according to the measurement data. Thus, the actual measurement data of the elbow is calculated through the vision system, and whether the elbow is a qualified product is determined according to the actual measurement data, so that automatic classification and screening of the elbow is realized, the efficiency of automatic production of air conditioners is improved, defective elbows are prevented from being installed on the air conditioners, and high product quality of the air conditioners is ensured.

[0180] According to the embodiment of the present application, a computer program product corresponding to the screening method of the elbow is also provided, which comprises a computer program, and the computer program product is processed to realize the steps of the above-mentioned screening method of the elbow.

[0181] Since the processing and functions realized by the computer program product of the present embodiment are basically corresponding to the above-mentioned embodiments, principles and examples of the method, the descriptions not described in detail in the present embodiment can be referred to the related descriptions in the above-mentioned embodiments, which will not be described here.

[0182] The technical scheme of the present application is provided with a vision system, which comprises a first camera and a second camera; the first camera is used for collecting a side view image of the elbow; the second camera is used for collecting a top view image of the elbow; measurement data of the elbow is determined according to the side view image and the top view image, and the elbow is determined as a qualified product or a defective product according to the measurement data. Thus, the actual measurement data of the elbow is calculated through the vision system, and whether the elbow is a qualified product is determined according to the actual measurement data, so that automatic classification and screening of the elbow is realized, the efficiency of automatic production of air conditioners is improved, defective elbows are prevented from being installed on the air conditioners, and high product quality of the air conditioners is ensured.

[0183] In conclusion, those skilled in the art can easily understand that the above-mentioned advantageous modes can be freely combined and superimposed without conflict.

[0184] The above-mentioned only describes the embodiments of the present application, and is not used to limit the present application. Those skilled in the art can make various modifications and changes to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the scope of claims of the present application.​​​​​

Claims

1. A method of screening for elbows, characterized in that, A vision system is provided, which includes: a first camera, a second camera; the first camera is used to collect a side view image of the elbow; the second camera is used to collect a top view image of the elbow; The method comprises: acquiring the side view image of the elbow collected by the first camera and the top view image of the elbow collected by the second camera; determining measurement data of the elbow according to the side view image and the top view image; the measurement data includes: a height of the elbow, a girth welding height of the elbow, a pipe opening diameter of the elbow, and a pipe opening center distance of the elbow; determining whether the elbow is a qualified product or a defective product according to the measurement data; wherein determining the measurement data of the elbow according to the side view image and the top view image comprises: determining the height of the elbow and the girth welding height of the elbow according to the side view image, and determining the pipe opening diameter of the elbow and the pipe opening center distance of the elbow according to the top view image; determining the height of the elbow and the girth welding height of the elbow according to the side view image comprises: correcting distortion of the side view image; locating features and positions of the elbow in the side view image, and correcting the positions of the elbow by using preset correction parameters; framing a bottom ROI area of the elbow to determine pixel coordinates of a reference straight line of the bottom of the elbow; framing a top ROI area of the elbow to determine pixel coordinates of a vertex of the elbow; respectively framing a left girth welding ROI area and a right girth welding ROI area of the elbow to determine pixel coordinates of a first point and a second point of the left girth welding, and pixel coordinates of a first point and a second point of the right girth welding; inputting the pixel coordinates of the vertex of the elbow and the pixel coordinates of the reference straight line of the bottom of the elbow into a preset elbow height calculation formula to obtain pixel coordinates of the height of the elbow; inputting the pixel coordinates of the first point and the second point of the left girth welding, the pixel coordinates of the first point and the second point of the right girth welding, and the pixel coordinates of the reference straight line of the bottom of the elbow into a preset girth welding height calculation formula to obtain pixel coordinates of the girth welding height of the elbow; respectively converting the pixel coordinates of the height of the elbow and the pixel coordinates of the girth welding height of the elbow into world coordinates to obtain the height of the elbow and the girth welding height of the elbow.

2. The method of claim 1, wherein, determining the pipe opening diameter of the elbow and the pipe opening center distance of the elbow according to the top view image comprises: correcting distortion of the top view image; locating features and positions of the elbow in the top view image, and correcting the positions of the elbow by using preset correction parameters; respectively framing inner circle and outer circle ROI areas of a first pipe opening and a second pipe opening of the elbow to determine pixel coordinates of a first pipe opening center, pixel coordinates of a second pipe opening center, and pixel coordinates of a pipe opening diameter of the elbow; determining pixel coordinates of the pipe opening center distance of the elbow according to the pixel coordinates of the first pipe opening center and the pixel coordinates of the second pipe opening center; respectively, the pixel coordinates of the pipe orifice center distance of the elbow are converted into world coordinates to obtain the pipe orifice diameter of the elbow and the pipe orifice center distance of the elbow.

3. The method of claim 1, wherein, According to the measurement data, it is determined whether the elbow is a qualified product or a substandard product, comprising: respectively, the pixel coordinates of the pipe orifice center distance of the elbow are converted into world coordinates to obtain the pipe orifice diameter of the elbow and the pipe orifice center distance of the elbow. If the absolute value of the difference between the height of the elbow and the preset elbow height value is less than a preset first tolerance, and the absolute value of the difference between the girth welding height of the elbow and the preset girth welding height value is less than a preset second tolerance, and the absolute value of the difference between the pipe orifice diameter of the elbow and the preset pipe orifice diameter value is less than a preset third tolerance, and the absolute value of the difference between the pipe orifice center distance of the elbow and the preset pipe orifice center distance value is less than a preset fourth tolerance, it is determined that the elbow is a qualified product. If the absolute value of the difference between the height of the elbow and the preset elbow height value is greater than or equal to the preset first tolerance, or the absolute value of the difference between the girth welding height of the elbow and the preset girth welding height value is greater than or equal to the preset second tolerance, or the absolute value of the difference between the pipe orifice diameter of the elbow and the preset pipe orifice diameter value is greater than or equal to the preset third tolerance, or the absolute value of the difference between the pipe orifice center distance of the elbow and the preset pipe orifice center distance value is greater than or equal to the preset fourth tolerance, it is determined that the elbow is a substandard product.

4. The method of claim 3, wherein, Further comprising: After it is determined that the elbow is a qualified product, the preset elbow height value, the preset girth welding height value, the preset pipe orifice diameter value, and the preset pipe orifice center distance value are adjusted using the measurement data.

5. An apparatus for screening elbows, characterized by A visual system is provided, comprising: a first camera, a second camera; the first camera is used to collect the side view image of the elbow; the second camera is used to collect the overhead view image of the elbow; The device comprises: An acquisition unit is configured to acquire the side view image of the elbow collected by the first camera and the overhead view image of the elbow collected by the second camera; A determination unit is configured to determine the measurement data of the elbow according to the side view image and the overhead view image; the measurement data comprises: the height of the elbow, the girth welding height of the elbow, the pipe orifice diameter of the elbow, and the pipe orifice center distance of the elbow; The determination unit is further configured to determine whether the elbow is a qualified product or a substandard product according to the measurement data; The determination unit determines the measurement data of the elbow according to the side view image and the overhead view image, comprising: The height of the elbow and the girth welding height of the elbow are determined according to the side view image, and the pipe orifice diameter of the elbow and the pipe orifice center distance of the elbow are determined according to the overhead view image; The determination unit determines the height of the elbow and the girth welding height of the elbow according to the side view image, comprising: The side view image is corrected for distortion; In the side view image, the features and positions of the elbow are located, and the positions of the elbow are corrected using preset correction parameters; The bottom ROI region of the elbow is framed, and the pixel coordinates of the reference straight line of the bottom of the elbow are determined; The top ROI region of the elbow is framed, and the pixel coordinates of the vertex of the elbow are determined; The left and right girth welding ROI regions of the elbow are respectively framed, and the pixel coordinates of the first and second point positions of the left girth welding and the pixel coordinates of the first and second point positions of the right girth welding are determined; The pixel coordinates of the vertex of the elbow and the pixel coordinates of the reference straight line of the bottom of the elbow are brought into a preset elbow height calculation formula to obtain the pixel coordinates of the height of the elbow; The pixel coordinates of the first and second point positions of the left girth welding, the pixel coordinates of the first and second point positions of the right girth welding, and the pixel coordinates of the reference straight line of the bottom of the elbow are brought into a preset girth welding height calculation formula to obtain the pixel coordinates of the girth welding height of the elbow; The pixel coordinates of the height of the elbow and the pixel coordinates of the girth welding height of the elbow are respectively converted into world coordinates to obtain the height of the elbow and the girth welding height of the elbow.

6. An elbow screening and sorting system characterized by, Comprise: The elbow screening device of claim 5.

7. A storage medium, characterized by The storage medium comprises a stored program, wherein the program controls the device where the storage medium is located to execute the elbow screening method of any one of claims 1 to 4 when the program is running.

8. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the method of any one of claims 1 to 4.

Images