CN207472194U - A kind of box for material circulation volume weight measuring system - Google Patents

Abstract

The utility model discloses a kind of box for material circulation volume weight measuring systems, including mounting bracket and the conveyer belt being arranged on the mounting bracket, electronic scale, buphthalmos ball plate, fine-tuning stent, signal processing system, X-direction trimming part, Y-direction trimming part, Z-direction trimming part, LED bar graph light source, industrial camera and laser ranging module；The area data and input information that altitude information that signal processing system is obtained according to laser ranging module, image analysis obtain calculate box for material circulation volume, while provide its weight；The utility model can be realized while carry out real-time automatic measuring, efficiently and accurately to the volume and weight of box for material circulation on assembly line；Need not carry out Feature Points Matching compared to light curtain method or bis- (more) mesh vision measurement systems, this system, avoid the low texture of object, repeat texture, overlapping and it is discontinuous the problems such as, and substantially reduce computation complexity.

Description

A logistics box volume and weight measurement system

technical field

The utility model relates to the technical field of non-contact intelligent measurement and control, in particular to a system for measuring the volume and weight of a logistics box.

Background technique

Visual measurement technology belongs to computer vision technology. It is an engineering discipline that takes pictures of objects through cameras, obtains image information, and then performs image analysis to obtain information such as the pose and size of objects. It is divided into monocular vision and double (multi) vision. Vision technology; for measuring objects, computer vision is not affected by artificial factors, and its accuracy is not limited by the accuracy of reference objects such as measuring scales. Compared with traditional manual scale measurement, it has the advantages of fast, accurate, non-contact and low cost. It is suitable for batch automated measurement, which greatly reduces labor costs and improves production efficiency.

In the monocular vision technology, there is no need for feature point matching and parallax calculation, which reduces the amount of system calculation and improves the system operation speed and system stability. However, monocular vision can only obtain the position information of planar objects, and cannot measure its depth information.

In binocular or multi-eye vision technology, it is necessary to perform feature point matching and disparity calculation on objects. This process is affected by optical distortion and noise, specular reflection on smooth surfaces, projection reduction, perspective distortion, low texture, repeated texture, image overlap and non-linearity. The interference of continuous and other problems greatly reduces the matching accuracy and system stability, and the matching process has a large amount of computation.

In addition, computer vision often uses image segmentation processing methods, and automatic threshold segmentation methods such as fixed threshold segmentation, edge segmentation, and maximum between-class variance can be used in the segmentation process. The fixed threshold segmentation method is seriously affected by illumination changes; the edge segmentation method is seriously disturbed by object texture; the maximum inter-class variance method can effectively suppress the impact of illumination changes, but if the mirror reflection occurs on the smooth surface of the object, or the gray surface of the object surface and the background There are obvious 3 or more gray levels, which often lead to only segmenting the area of reflection or the highest brightness gray level, while ignoring other parts of the object.

Utility model content

The purpose of the utility model is to overcome the shortcomings and deficiencies of the prior art, and provide a volume and weight measurement system for logistics boxes.

The purpose of this utility model is achieved through the following technical solutions:

A logistics box volumetric weight measurement system, including a mounting frame, and a conveyor belt arranged on the mounting frame, an electronic scale, a bull's-eye ball plate, a fine-tuning bracket, a signal processing system, an X-direction fine-tuning component, a Y-direction fine-tuning component, a Z-direction fine-tuning component, a Direction fine-tuning components, LED strip light sources, industrial cameras and laser distance measuring modules, wherein the conveyor belt is arranged on the left and right sides of the bull's-eye ball board, and the electronic scale is arranged under the bull's-eye ball board; The fine-tuning parts in the X direction, the fine-tuning parts in the Y direction, the fine-tuning parts in the Z direction, the LED bar light source, the industrial camera and the laser ranging module are all arranged on the fine-tuning bracket, and the fine-tuning parts in the X-direction and the fine-tuning parts in the Y direction , LED strip light source, industrial camera and laser ranging module are all set on the top of the fine-tuning bracket, the industrial camera and laser ranging module are arranged side by side and are located directly above the bull’s-eye ball plate, and the LED strip The light source is arranged around the industrial camera and the laser ranging module and is located below the industrial camera and the laser ranging module; the X direction fine-tuning component controls the left and right movement of the industrial camera and the laser ranging module, and the Y direction fine-tuning The component controls the industrial camera and the laser ranging module to move back and forth, and the Z direction fine-tuning component controls the industrial camera and the laser ranging module to move up and down;

The industrial camera, the laser ranging module and the electronic scale are respectively connected with the signal processing system; the signal processing system is provided with a display screen.

Preferably, the bull's-eye ball plate is a black or dark metal bottom plate.

Preferably, the bull's-eye balls on the bull's-eye ball plate are arranged in an array.

Working principle of the utility model:

When working, step 1, start the logistics box volume and weight measurement system, industrial camera, laser distance measuring module, LED bar light source, electronic scale and signal processing system start to work, adjust the fine-tuning parts in the X direction, the fine-tuning parts in the Y direction and the fine-tuning in the Z direction part;

Step 2. Calibrate the internal parameters, external parameters and distortion coefficient of the industrial camera, as well as the distance between the laser ranging module and the highest point of the bull’s-eye ball plate, and input the above parameters into the signal processing system. The specific workflow is as follows:

(1) Camera calibration:

The industrial camera satisfies the pinhole camera model, and the image coordinate vector is Where (u, v) is the pixel coordinates of the target point; the internal parameter matrix of the camera is Where f _x , f _y , c _x , c _y are the x-direction focal length, y-direction focal length and optical axis center coordinates respectively; the camera extrinsic parameter matrix is (R|T), where R is the camera optical center coordinate system relative to the world coordinate system The 3×3 rotation matrix, T is the 3×1 translation matrix of the camera optical center coordinate system relative to the world coordinate system; the world coordinate vector is Among them, X, Y, and Z are the world coordinates of the target point; the camera imaging model satisfies the relation:

where z _c is the scale factor; in addition, the camera distortion model satisfies the following relationship:

Among them, (x, y) are the physical coordinates of the image before distortion correction, (x _cor , y _cor ) are the physical coordinates of the image after distortion correction, r=x ² +y ² ,k ₁ ,k ₂ ,k ₃ ,p ₁ , p ₂ are three radial distortion coefficients and two tangential distortion coefficients of the camera;

The purpose of camera calibration is to solve the internal parameters, external parameters and distortion coefficient of the camera, set the plane where the highest point of the bull’s-eye ball board is located as the zero plane, and adopt Zhang Zhengyou’s checkerboard plane calibration method to take 20 images by changing the position and angle of the checkerboard. Carry out camera calibration, and solve to obtain camera internal parameters, external parameters and distortion coefficients;

(2) Laser ranging module calibration:

Place the chessboard on the bull's-eye ball board, measure the distance between the board and the laser ranging module through the laser ranging module, and then compensate the thickness of the chessboard to obtain the distance between the laser ranging module and the highest point of the bull's-eye ball board;

Step 3, select the ROI (Region Of Interest) area, and shield the area outside the bull's-eye ball plate in the camera's field of view;

Step 4, put the logistics box into the conveyor belt, the logistics box enters the bull's-eye ball plate through the conveyor belt, and the electronic scale detects the weight of the logistics box. When the data of the electronic scale is stable, record the weight of the logistics box and transmit the data to the signal processing system, then Start the laser ranging module to obtain the height H of the logistics box;

Step 5: Start the logistics box volume identification algorithm process based on the nested loop maximum variance between classes. The specific workflow is as follows:

(1) Input camera intrinsic parameters f _x , camera extrinsic parameters T _z , camera distortion coefficient, logistics box height H and logistics box image G ₀ ;

(2) If the height H of the logistics box is less than the preset threshold, an error message is returned to the signal processing system, and step six is started; otherwise, the distortion coefficient of the camera is used to perform de-distortion processing on the logistics box image to obtain the de-distorted image G;

(3) Use the method of maximum variance between classes to carry out the first binarization process on the logistics box image data, and the target area in the binary image B ₀ obtained by segmentation is the highlighted area, and then define the intermediate binary image B _tmp =B ₀ ;

(4) Invert the binary image B _tmp to obtain the mask image M;

(5) For the region in the dedistorted image G corresponding to the highlighted region of the mask image M, the binarization process of the maximum inter-class variance method is performed to obtain the inter-class variance D and the binary image of the highlighted target, and Assign the binary image of the highlighted target to B _tmp , if the inter-class variance D is greater than the preset threshold, add the highlighted area of B _tmp to the binary image B ₀ , and return to step 5 (4), otherwise proceed to the next step;

(6) Find the maximum contour of the binary image B ₀ , calculate and fill its convex hull contour C with highlighted gray scale, and calculate the pixel length L, pixel width W and vertex pixel coordinates P of the smallest circumscribed rectangle of the convex hull contour C= {P ₁ ,P ₂ ,P ₃ ,P ₄ };

(7) Carry out logistics box out-of-boundary detection and interference detection, the specific workflow is as follows:

① Input the ROI information obtained in step 3 and the C, L, W and P obtained in step 5 (6);

② Use the boundary information of ROI to judge and record the out-of-bounds elements in the vertex pixel coordinates P;

③ If the number N of out-of-bounds vertices is greater than 1, and more than two adjacent vertices are out-of-bounds, return an error message to the signal processing system and start step 6; otherwise, according to the coordinates of the out-of-bounds vertex and its two adjacent vertices and the ROI Boundary information, using similar triangle properties to calculate the area of the out-of-bounds area, the area of the out-of-bounds area is which is The same is true for other boundary out-of-boundary situations, and the total out-of-boundary area is If no vertex is out of bounds, then

④ Carry out interference detection, calculate the area S _C of the convex hull contour C, if the area ratio If it is less than the preset threshold, an error message is returned to the signal processing system, and step six is started; otherwise, step five (8) is started;

(8) If the out-of-bounds detection and interference detection of the logistics box in step 5 (7) pass, calculate the volume of the logistics box And return the logistics box volume information to the signal processing system;

Step 6, the display screen of the signal processing system displays the volume and weight information of the logistics box, or displays a prompt message that the logistics box is out of bounds. .

Compared with the prior art, the utility model has the following beneficial effects:

(1) Compared with the traditional manual scale measurement, the utility model has the advantages of fast, accurate, non-contact and low cost, and is suitable for batch automatic measurement, which greatly reduces labor costs and improves production efficiency;

(2) Compared with the binocular or multi-eye vision system, the utility model adopts the monocular vision system, which does not need to perform feature point matching and parallax calculation, which reduces the amount of system calculation, improves the system operation speed and system stability, and at the same time The utility model introduces a laser ranging module to obtain the depth information of the logistics box in the image, and realizes the measurement of the volume of the logistics box efficiently. Furthermore, compared with the laser ranging trigger method that is easy to cause stable ranging but unstable weight, The utility model adopts weight stable trigger, which is beneficial to obtain stable and non-interfering weight and volume data;

(3) The image segmentation method of the nested loop maximum variance between classes adopted by the utility model, compared with the image segmentation methods of the fixed threshold segmentation method, the edge segmentation method and the maximum variance between the clusters, the utility model can effectively weaken the illumination change, The effect of object surface texture and smooth surface specular reflection on segmentation;

(4) The utility model has carried out out-of-bounds and interference detection on the target, effectively judging whether the target is out of bounds, and whether there are other interference objects in the camera field of view that cause measurement errors, and the out-of-bounds and interference detection can effectively filter the measurement data, It greatly reduces the risk of wrong data being mistaken for correct data, and gives specific out-of-bounds or interference prompt information, so that operators can quickly adjust the logistics box or system.

Description of drawings

Fig. 1 is the structural representation of the utility model;

Fig. 2 is the overall flowchart of the system recognition algorithm part of the present utility model;

Fig. 3 is the flow chart of the logistics box volume recognition algorithm based on the maximum variance between classes of nested loops of the present invention;

Fig. 4 is a flow chart of out-of-bounds and interference detection algorithm of the present utility model;

Fig. 5 is a schematic diagram of the calculation method of the right boundary out of the present invention.

The reference signs in the figure are: 1. Signal processing system; 2. Industrial camera; 3. Laser ranging module; 4. LED bar light source; 5. Bull's eye ball board; 6. Electronic scale; 7. Conveyor belt; Fine-tuning bracket; 9. Z-direction fine-tuning part; 10. X-direction fine-tuning part; 11. Y-direction fine-tuning part; 12. Mounting frame.

Detailed ways

The utility model will be further described in detail below in conjunction with the embodiments and accompanying drawings, but the implementation of the utility model is not limited thereto.

As shown in Figures 1 to 5, a logistics box volumetric weight measurement system includes a mounting frame 12, and a conveyor belt 7, an electronic scale 6, a bull's-eye ball plate 5, a fine-tuning bracket 8, and a signal mounted on the mounting frame 12. Processing system 1, X-direction fine-tuning part 10, Y-direction fine-tuning part 11, Z-direction fine-tuning part 9, LED bar light source 4, industrial camera 2 and laser distance measuring module 3, wherein, the bull's-eye ball plate 5 is black or The dark-toned metal bottom plate can facilitate the collection of images in this way; at the same time, the bull’s-eye balls on the bull’s-eye ball plate 5 are arranged in an array, so that the logistics box can be conveniently entered and exited by the bull’s-eye ball plate 5; the conveyor belt 7 It is arranged on the left and right sides of the bull's-eye ball plate 5, and the electronic scale 6 is arranged under the bull's-eye ball plate 5; , LED bar-shaped light source 4, industrial camera 2 and laser ranging module 3 are all arranged on the fine-tuning bracket 8, and the X-direction fine-tuning part 10, Y-direction fine-tuning part 11, LED bar-shaped light source 4, industrial camera 2 and the laser distance measuring module 3 are all arranged on the top of the fine-tuning bracket 8, the industrial camera 2 and the laser distance measuring module 3 are arranged side by side and are located directly above the bull's-eye ball plate 5, and the LED bar light source 4 is arranged Around the industrial camera 2 and the laser rangefinder module 3 and below the industrial camera 2 and the laser rangefinder module 3, an LED strip light source 4 can be set to achieve uniform lighting for the industrial camera 2 to obtain high-quality images; X-direction fine-tuning part 10 controls described industrial camera 2 and laser ranging module 3 to move left and right, described Y-direction fine-tuning part 11 controls described industrial camera 2 and laser distance-finding module 3 to move forward and backward, and described Z-direction fine-tuning part 9 controls The industrial camera 2 and the laser distance measuring module 3 move up and down; the industrial camera 2, the laser distance measuring module 3 and the electronic scale 6 are respectively connected with the signal processing system 1; the signal processing system 1 is provided with a display screen .

The utility model provides an image segmentation method with a nested loop maximum variance between classes, and introduces a laser ranging module 3 for obtaining object depth information, and combines the object two-dimensional information obtained by monocular vision technology to realize logistics Accurate non-contact measurement of the length, width and height of the box, while using the traditional electronic scale 6 to generate trigger and weighing functions, and finally realize the convenient, automatic and high-accuracy measurement of the volume and weight of the logistics box.

When working, step 1 starts the logistics box volume and weight measurement system, the industrial camera 2, the laser distance measuring module 3, the LED bar light source 4, the electronic scale 6 and the signal processing system 1 start to work, and adjust the fine-tuning parts 10 in the X direction and the Y direction A fine-tuning component 11 and a Z-direction fine-tuning component 9;

Step 2: Carry out the calibration of the internal parameters, external parameters and distortion coefficient of the industrial camera 2, as well as the calibration of the distance between the laser ranging module 3 and the highest point of the bull’s-eye ball plate 5, and input the above parameters into the signal processing system 1, and the specific work The process is as follows:

(1) Camera calibration:

The industrial camera 2 satisfies the pinhole camera model, and the image coordinate vector is set as Where (u, v) is the pixel coordinates of the target point; the internal parameter matrix of the camera is Where f _x , f _y , c _x , c _y are the x-direction focal length, y-direction focal length and optical axis center coordinates respectively; the camera extrinsic parameter matrix is (R|T), where R is the camera optical center coordinate system relative to the world coordinate system The 3×3 rotation matrix, T is the 3×1 translation matrix of the camera optical center coordinate system relative to the world coordinate system; the world coordinate vector is Among them, X, Y, and Z are the world coordinates of the target point; the camera imaging model satisfies the relation:

where z _c is the scale factor; in addition, the camera distortion model satisfies the following relationship:

Among them, (x, y) are the physical coordinates of the image before distortion correction, (x _cor , y _cor ) are the physical coordinates of the image after distortion correction, r=x ² +y ² ,k ₁ ,k ₂ ,k ₃ ,p ₁ , p ₂ are three radial distortion coefficients and two tangential distortion coefficients of the camera;

The purpose of camera calibration is to solve the internal parameters, external parameters and distortion coefficient of the camera, set the plane at the highest point of the bull’s-eye ball board 5 as the zero plane, and adopt Zhang Zhengyou’s checkerboard plane calibration method to take 20 images by changing the position and angle of the checkerboard , carry out camera calibration, and solve to obtain camera intrinsic parameters, extrinsic parameters and distortion coefficients;

(2) Laser ranging module 3 calibration:

Place the checkerboard on the bull's-eye ball board 5, measure the distance between the checkerboard and the laser range-finding module 3 through the laser range-finding module 3, and then compensate the thickness of the checkerboard to obtain the highest distance between the laser range-finding module 3 and the bull's-eye ball board 5. distance between points;

Step 3, select the ROI (Region Of Interest) area, and shield the areas other than the bull's-eye ball plate 5 in the field of view of the camera;

Step 4, put the logistics box into the conveyor belt 7, the logistics box enters the bull's-eye ball plate 5 through the conveyor belt 7, and the electronic scale 6 detects the weight of the logistics box. When the data of the electronic scale 6 is stable, record the weight of the logistics box and transmit the data to The signal processing system 1 starts the laser ranging module 3 afterwards to obtain the height H of the logistics box;

Step 5: Start the logistics box volume identification algorithm process based on the nested loop maximum variance between classes, as shown in Figure 3, the specific workflow is as follows:

(1) Input camera intrinsic parameters f _x , camera extrinsic parameters T _z , camera distortion coefficient, logistics box height H and logistics box image G ₀ ;

(2) If the height H of the logistics box is less than the preset threshold (the preset threshold here is the empirical value obtained by summarizing the experimental data after many experiments), then return an error message to the signal processing system 1, and start step 6; otherwise Use the distortion coefficient of the camera to perform de-distortion processing on the logistics box image to obtain the de-distorted image G;

(3) Use the method of maximum variance between classes to carry out the first binarization process on the logistics box image data, and the target area in the binary image B ₀ obtained by segmentation is the highlighted area, and then define the intermediate binary image B _tmp =B ₀ ;

(4) Invert the binary image B _tmp to obtain the mask image M;

(5) For the region in the dedistorted image G corresponding to the highlighted region of the mask image M, the binarization process of the maximum inter-class variance method is performed to obtain the inter-class variance D and the binary image of the highlighted target, and Get the binary image of the highlighted target and assign it to _Btmp . If the inter-class variance D is greater than the preset threshold (the preset threshold here is the empirical value obtained by summarizing the experimental data after many experiments), then the high value of _Btmp Add the bright area to the binary image B ₀ , and get back to step five (4), otherwise proceed to the next step;

(6) Find the maximum contour of the binary image B ₀ , calculate and fill its convex hull contour C with highlighted gray scale, and calculate the pixel length L, pixel width W and vertex pixel coordinates P of the smallest circumscribed rectangle of the convex hull contour C= {P ₁ ,P ₂ ,P ₃ ,P ₄ };

(7) Carry out logistics box out-of-bounds detection and interference detection, as shown in Figure 4, the specific workflow is as follows:

① Input the ROI information obtained in step 3 and the C, L, W and P obtained in step 5 (6);

② Use the boundary information of ROI to judge and record the out-of-bounds elements in the vertex pixel coordinates P;

③ If the number N of out-of-bound vertices is greater than 1, and more than two adjacent vertices are out-of-bound, return an error message to signal processing system 1, and start step six; otherwise, according to the coordinate sum of the out-of-bound vertex and its two adjacent vertices, For ROI boundary information, use the similar triangle property to calculate the area of the out-of-bounds area, as shown in Figure 5, the calculation method for the right-boundary out-of-bounds area, and the area of the out-of-bounds area is which is The same is true for other boundary out-of-boundary situations, and the total out-of-boundary area is If no vertex is out of bounds, then

④ Carry out interference detection, calculate the area S _C of the convex hull contour C, if the area ratio Less than the preset threshold (the preset threshold here is the empirical value obtained by summarizing the experimental data after many experiments), then return error information to the signal processing system 1, and start step six; otherwise start step five (8);

(8) If the out-of-bounds detection and interference detection of the logistics box in step 5 (7) pass, calculate the volume of the logistics box And return the logistics box volume information to the signal processing system 1;

Step 6, the display screen of the signal processing system 1 displays the volume and weight information of the logistics box, or displays a prompt message that the logistics box is out of bounds.

Compared with the traditional manual ruler measurement, the utility model has the advantages of fast, accurate, non-contact and low cost, and is suitable for batch automatic measurement, which greatly reduces labor costs and improves production efficiency; compared with binocular or multi-eye Vision system, the utility model adopts a monocular vision system, which does not need to perform feature point matching and parallax calculation, which reduces the amount of calculation of the system, improves the system operation speed and system stability, and at the same time, the utility model introduces a laser ranging module to acquire images The depth information of the logistics box in the logistics box efficiently realizes the measurement of the volume of the logistics box. Furthermore, compared with the laser ranging trigger method that is easy to cause stable distance measurement but unstable weight, the utility model uses stable weight triggering, which is beneficial to obtain stable weight. and non-interfering weight and volume data; the image segmentation method of the nested loop maximum variance between classes adopted, compared with the image segmentation methods of fixed threshold segmentation method, edge segmentation method and maximum variance between classes, the utility model can effectively reduce the The impact of illumination changes, object surface texture, and smooth surface specular reflection on segmentation; the out-of-bounds and interference detection of the target is carried out to effectively determine whether the target is out of bounds, and whether there are other interference objects in the camera's field of view that cause measurement errors, the out-of-bounds and Interference detection effectively filters the measurement data, which greatly reduces the risk of mistaking wrong data as correct data, and gives specific out-of-bounds or interference prompt information, allowing operators to quickly adjust logistics boxes or systems.

The above is a preferred implementation of the utility model, but the implementation of the utility model is not limited by the above content, and any other changes, modifications, substitutions, combinations, Simplification should be equivalent replacement methods, all of which are included in the protection scope of the present utility model.

Claims (3)

1. A logistics box volumetric weight measurement system, characterized in that it includes a mounting frame, and a conveyor belt, an electronic scale, a bull's-eye ball plate, a fine-tuning bracket, a signal processing system, a fine-tuning component in the X direction, and a conveyor belt arranged on the mounting frame. Fine-tuning parts in the Y direction, fine-tuning parts in the Z direction, LED bar light sources, industrial cameras and laser distance measuring modules, wherein the conveyor belt is set on the left and right sides of the bull's-eye ball board, and the electronic scale is set on the bull's-eye Below the eye ball plate; the fine-tuning parts in the X direction, the fine-tuning parts in the Y direction, the fine-tuning parts in the Z direction, the LED bar light source, the industrial camera and the laser ranging module are all arranged on the fine-tuning bracket, and the fine-tuning parts in the X direction Components, Y-direction fine-tuning components, LED bar light source, industrial camera and laser ranging module are all arranged on the top of the fine-tuning bracket, and the industrial camera and laser ranging module are arranged side by side and located directly above the bull’s-eye ball plate , the LED bar-shaped light source is arranged around and below the industrial camera and the laser ranging module; the X direction fine-tuning component controls the left and right movement of the industrial camera and the laser ranging module , the Y-direction fine-tuning component controls the industrial camera and the laser ranging module to move back and forth, and the Z-direction fine-tuning component controls the industrial camera and the laser ranging module to move up and down; The industrial camera, the laser ranging module and the electronic scale are respectively connected with the signal processing system; the signal processing system is provided with a display screen. 2. The logistics box volumetric weight measuring system according to claim 1, characterized in that, the bull's-eye ball plate is a black or dark metal bottom plate. 3. The logistics box volumetric weight measuring system according to claim 1, wherein the bull's-eye balls on the bull's-eye ball plate are arranged in an array.