阅读说明：本技术 一种基于视觉的汽车轮毂定位方法 (Vision-based automobile hub positioning method ) 是由 方灶军 李俊杰 李华民 谌鸿强 杨桂林 张驰 于 2018-05-25 设计创作，主要内容包括：本发明公开了一种基于视觉的汽车轮毂定位方法,在机器人粗抛工位和机器人精抛工位之前增加视觉定位工位；在视觉定位工位通过对轮毂视觉取像、轮毂中心点计算、轮毂条幅角度计算,得到当前轮毂条幅的角度α<Sub>f</Sub>；根据当前轮毂条幅角度α<Sub>f</Sub>与轮毂条幅基准角度α<Sub>r</Sub>的差值,通过放置轮毂的转台进行差值调整,将轮毂角度调整到设定的基准位置。(The invention discloses a vision-based automobile hub positioning method, which is characterized in that a vision positioning station is added before a robot rough polishing station and a robot fine polishing station; obtaining the angle alpha of the current hub banner through the hub vision image capture, the hub central point calculation and the hub banner angle calculation at the vision positioning station f (ii) a According to the current hub scroll angle alpha f Reference angle alpha with hub spoke r The difference value is adjusted through a rotary table for placing the wheel hub, and the wheel hub angle is adjusted to a set reference position.)

1. A vision-based automobile hub positioning method is characterized in that: a vision positioning station is added before a robot rough polishing station and a robot fine polishing station; obtaining the angle alpha of the current hub banner through the hub vision image capture, the hub central point calculation and the hub banner angle calculation at the vision positioning station_f(ii) a According to the current hub scroll angle alpha_fReference angle alpha with hub spoke_rThe difference value is adjusted through a rotary table for placing the wheel hub, and the current wheel hub angle is adjusted to a set reference position.

2. The vision-based automobile hub positioning method according to claim 1, characterized in that: will be used for measuring the reference angle alpha of the scroll_rThe wheel hub is placed on the rotary table, and the rotary table is manually rotated to enable the initial polishing point of the wheel hub to be aligned with the polishing action starting point of the robot; then, using image processing software to perform visual image capture, hub central point calculation and hub spoke angle calculation on the hub aligned with the base point to obtain a hub spoke reference angle alpha_rAnd the reference angle is used as the reference angle for the subsequent automatic adjustment of the hub angle.

3. The vision-based automobile hub positioning method according to claim 2, characterized in that: after the subsequent wheel hub is placed on the rotary table, the image processing software is used for carrying out visual image capture, wheel hub central point calculation and wheel hub spoke angle calculation on the wheel hub to obtain a wheel hub spoke angle alpha_f(ii) a Calculating the current hub spoke angle alpha_fReference angle alpha with hub spoke_rA difference of (a), i.e. a ═ a_f-α_r(ii) a And sending the wheel hub spoke angle difference value delta alpha to a wheel hub turntable controller, and driving the wheel hub to automatically adjust the angle by the turntable so as to enable the angle of the subsequent wheel hub to reach a reference angle.

4. The vision-based automobile hub positioning method according to claim 1, characterized in that: the hub center point calculation comprises extracting by using a canny operatorThe edge in the hub image is taken out, the hub center circular hole in the image is detected by the Hough transformation algorithm, and the circle center (x) detected by the Hough transformation algorithm is used_0w,y_0w) As the center of the hub.

5. The vision-based automobile hub positioning method according to claim 4, characterized in that: the main steps of detecting the central circular hole of the hub in the image through the Hough transformation algorithm are as follows:

step a1, writing the circle in polar form, i.e.Wherein x and y are sample points on a circle, x₀、y₀As the coordinate of the center of the circle, r is the radius of the circle, and theta is the polar angle;

step a2, pair (x)₀,y₀) And theta is sampled according to a certain step length, and the sample point (x) on the edge is sampled_i,y_i) Substituting into the polar equation of circle to obtain the corresponding r value (x)_i,y_i)∈B_s，(i＝1,2,3,…N_p) In which B is_sIs a set of image edge points, N_pIs the number of edge points;

step a3, eliminating unreasonable data pairs (x) by utilizing the value range of r_0j,y_0j,r_j)，r_j＜r_minOr r_j＞r_maxWherein r is_minAnd r_maxThe radius of the central circular hole of the hub is the minimum value and the maximum value;

step a4, for the remaining data pairs (x)_0k,y_0k,r_k) Voting is carried out, the data pair with the largest number of votes is the extracted circle parameter and is marked as (x)_0w,y_0w,r_w)；

Step a5, circle center (x) detected by Hough transformation algorithm_0w,y_0w) As the center of the hub.

6. The vision-based automobile hub positioning method according to claim 1, characterized in that: the calculation of the spoke angle of the hub comprises the following steps:

step b1, drawing a vertical line from top to bottom in the hub spoke length direction 1/3 to 2/3 as a reference line L, wherein L can be described as: x ═ x_l；

Step b2, calculating the upper edge P of the scroll along the reference line L₁And a lower edge point P₂(ii) a Upper edge point P₁The coordinate calculation formula isWherein (p)_1x,p_1y) Is the upper edge point P of the hub₁I (x, y) is the gray value of the pixel point (x, y) in the image; correspondingly, lower edge point P₂The coordinate calculation formula isWherein (p)_2x,p_2y) Is the lower edge point P of the hub₂The coordinates of (a);

step b3, passing the upper edge point P of the hub spoke₁And a lower edge point P₂Calculating the spoke center point P of the hub₀Coordinate of (1), hub spoke center point P₀Is calculated by the formulaWherein (p)_0x,p_0y) Is the central point P of the spoke of the hub₀The coordinates of (a);

step b4, passing through the hub central point O and the hub spoke central point P₀Calculating the angle alpha of the spoke of the hub according to the formula

Technical Field

The invention relates to the field of automobile hub polishing, in particular to an automobile hub positioning method based on vision.

Background

With the application of the robot hub polishing system, the full automation degree of the hub polishing technology is higher and higher. The robot hub polishing system generally comprises a robot rough polishing station and a robot fine polishing station. The rough polishing and the fine polishing of the robot are performed according to a fixed path, but the angle of each hub cannot be guaranteed to be completely consistent in the loading process, and the rough polishing and the fine polishing of the robot are seriously influenced.

Disclosure of Invention

The invention provides a vision-based automobile hub positioning method, which solves the problem of the consistency of the initial positions of hubs in a robot hub polishing system and provides accurate reference positioning for automatic repeated polishing of a robot.

The technical scheme adopted by the invention for solving the technical problems is as follows: a vision-based automobile hub positioning method is characterized in that a vision positioning station is added before a robot rough polishing station and a robot fine polishing station; obtaining the angle alpha of the current hub banner through the hub vision image capture, the hub central point calculation and the hub banner angle calculation at the vision positioning station_f(ii) a According to the current hub scroll angle alpha_fReference angle alpha with hub spoke_rThe difference value is adjusted through a rotary table for placing the wheel hub, and the wheel hub angle is adjusted to a set reference position.

In the above method, the reference angle alpha of the scroll is measured_rThe wheel hub is placed on the rotary table, and the rotary table is manually rotated to enable the initial polishing point of the wheel hub to be aligned with the polishing action starting point of the robot; then, using image processing software to perform visual image capture, hub central point calculation and hub spoke angle calculation on the hub aligned with the base point to obtain a hub spoke reference angle alpha_rAnd the reference angle is used as the reference angle for the subsequent automatic adjustment of the hub angle.

In the method, after the subsequent wheel hub is placed on the rotary table, the image processing software is used for carrying out visual image capture, wheel hub central point calculation and wheel hub spoke angle calculation on the wheel hub to obtain the wheel hub spoke angle alpha_f(ii) a Current hub swath angle alpha_fReference angle alpha with hub spoke_rA difference of (a), i.e. a ═ a_f-α_r(ii) a And sending the wheel hub spoke angle difference value delta alpha to a wheel hub turntable controller, and driving the wheel hub to automatically adjust the angle by the turntable so as to enable the angle of the subsequent wheel hub to reach a reference angle.

In the method, the calculation of the hub center point comprises the steps of extracting the edge in the hub image by adopting a canny operator, detecting the hub center circular hole in the image by adopting a Hough transformation algorithm, and detecting the circle center (x) detected by the Hough transformation algorithm_0w,y_0w) As the center of the hub.

In the method, the main steps of detecting the central circular hole of the hub in the image through the Hough transformation algorithm are as follows:

step a1, writing the circle in polar form, i.e.Wherein x and y are sample points on a circle, x₀、y₀As the coordinate of the center of the circle, r is the radius of the circle, and theta is the polar angle;

step a2, pair (x)₀,y₀) And theta is sampled according to a certain step length, and the sample point (x) on the edge is sampled_i,y_i) Substituting into the polar equation of circle to obtain the corresponding r value (x)_i,y_i)∈B_s，(i＝1,2,3,…N_p) In which B is_sIs a set of image edge points, N_pIs the number of edge points;

step a3, eliminating unreasonable data pairs (x) by utilizing the value range of r_0j,y_0j,r_j)，r_j＜r_minOr r_j＞r_maxWherein r is_minAnd r_maxThe radius of the central circular hole of the hub is the minimum value and the maximum value;

step a4, for the remaining data pairs (x)_0k,y_0k,r_k) Voting is carried out, the data pair with the largest number of votes is the extracted circle parameter and is marked as (x)_0w,y_0w,r_w)；

Step a5, circle center (x) detected by Hough transformation algorithm_0w,y_0w) As the center of the hub.

In the method, the calculation of the spoke angle of the hub comprises the following steps:

step b1, drawing a vertical line from top to bottom in the hub spoke length direction 1/3 to 2/3 as a reference line L, wherein L can be described as: x ═ x_l；

Step b2, calculating the upper edge P of the scroll along the reference line L₁And a lower edge point P₂；

Upper edge point P₁The coordinate calculation formula isWherein, (P1x, P1y) is the upper edge point P of the hub₁I (x, y) is the gray value of the pixel point (x, y) in the image; correspondingly, lower edge point P₂The coordinate calculation formula isWherein, (P2x, P2y) is the lower edge point P of the hub₂The coordinates of (a);

step b3, passing the upper edge point P of the hub spoke₁And a lower edge point P₂Calculating the spoke center point P of the hub₀Coordinate of (1), hub spoke center point P₀Is calculated by the formulaWherein, (P0x, P0y) is the central point P of the spoke of the hub₀The coordinates of (a);

step b4, passing through the hub central point O and the hub spoke central point P₀Calculating the angle alpha of the spoke of the hub according to the formula

Compared with the prior art, the vision-based automobile hub positioning method has the advantages that the non-contact vision positioning method is adopted to position the wheel hub, so that the positioning precision is improved; through visual positioning, the consistency of the initial positions of the hubs can be ensured, and accurate reference positions are provided for continuous and repeated polishing operation of the robot; the visual positioning algorithm consists of three parts, namely hub central point calculation, hub banner angle calculation and hub angle adjustment; the calculation of the central point of the hub is composed of edge extraction and Hough transformation, so that the precision and the reliability are guaranteed; in the wheel hub spoke angle calculation process, a special upper edge and lower edge extraction algorithm is designed according to the bright spot condition of the wheel hub spoke, and the accuracy and reliability of edge extraction are improved.

Drawings

Fig. 1 is a schematic view of a hub according to the present embodiment.

Fig. 2 is a flowchart of the method of the present embodiment.

Detailed Description

Embodiments of the present invention are described in further detail below with reference to the accompanying drawings.

Fig. 1 to 2 are schematic views of the present invention.

The vision-based automobile hub positioning method can realize automatic alignment in the hub feeding process, and facilitates repeated polishing operation of a robot according to a fixed track.

The method is characterized in that a visual positioning station is added before a robot rough polishing station and a robot fine polishing station. Obtaining the angle alpha f of the current hub banner at the visual positioning station through hub visual image capture, hub center point calculation and hub banner angle calculation; and according to the difference value between the current hub spoke angle alpha f and the hub spoke reference angle alpha r, performing difference value adjustment through a rotary table for placing the hub, and adjusting the current hub angle to a set reference position.

Will be used for measuring the reference angle alpha of the scroll_rThe wheel hub is placed on the rotary table, and the rotary table is manually rotated to enable the initial polishing point of the wheel hub to be aligned with the polishing action starting point of the robot; then, using image processing software to perform visual image capture, hub central point calculation and hub spoke angle calculation on the hub aligned with the base point to obtain a hub spoke reference angle alpha_rAnd the reference angle is used as the reference angle for the subsequent automatic adjustment of the hub angle.

After the subsequent wheel hub is placed on the rotary table, the image processing software is used for carrying out visual image capture, wheel hub central point calculation and wheel hub spoke angle calculation on the wheel hub to obtain a wheel hub spoke angle alpha_f(ii) a Current hub swath angle alpha_fWith hub spoke baseQuasi angle alpha_rA difference of (a), i.e. a ═ a_f-α_r(ii) a And sending the wheel hub spoke angle difference value delta alpha to a wheel hub turntable controller, and driving the wheel hub to automatically adjust the angle by the turntable so as to enable the angle of the subsequent wheel hub to reach a reference angle.

As shown in fig. 1, the automobile hub is in a schematic view, and the shape of the automobile hub is as shown in the figure and comprises a central circular hole 1, a scroll 2 and an outer frame 3. In the polishing and grinding process, the hub is fixed on the rotary table through the central circular hole 1, and the hub polishing process is completed by the coordination action of the robot and the rotary table.

Such as the method flow diagram shown in fig. 2. The calculation of the hub center point comprises the steps of extracting the edge in the hub image by adopting a canny operator, detecting a hub center round hole 1 in the image by adopting a Hough transformation algorithm, and detecting the circle center (x) detected by the Hough transformation algorithm_0w,y_0w) As the center of the hub.

The main steps of detecting the central circular hole 1 of the hub in the image through the Hough transformation algorithm are as follows:

step a1, writing the circle in polar form, i.e.Wherein x and y are sample points on a circle, x₀、y₀As the coordinate of the center of the circle, r is the radius of the circle, and theta is the polar angle;

step a2, pair (x)₀,y₀) And theta is sampled according to a certain step length, and the sample point (x) on the edge is sampled_i,y_i) Substituting into the polar equation of circle to obtain the corresponding r value (x)_i,y_i)∈B_s，(i＝1,2,3,…N_p) In which B is_sIs a set of image edge points, N_pIs the number of edge points;

step a3, eliminating unreasonable data pairs (x) by utilizing the value range of r_0j,y_0j,r_j)，r_j＜r_minOr r_j＞r_maxWherein r is_minAnd r_maxThe radius of the central circular hole 1 of the hub is the minimum value and the maximum value;

step a4, for the remaining data pairs (x)_0k,y_0k,r_k) Voting is carried out, the data pair with the largest number of votes is the extracted circle parameter and is marked as (x)_0w,y_0w,r_w)；

Step a5, circle center (x) detected by Hough transformation algorithm_0w,y_0w) As the center of the hub.

The calculation of the spoke angle of the hub comprises the following steps:

step b1, drawing a vertical line from top to bottom in the hub spoke length direction 1/3 to 2/3 as a reference line L, wherein L can be described as: x ═ x_l；

Step b2, calculating the upper edge P of the scroll along the reference line L₁And a lower edge point P₂(ii) a Because the hub spoke is made of aluminum alloy material, the brightness is higher under the irradiation of a light source, and the brightness of other areas is lower; accordingly, the upper edge point P₁The coordinate calculation formula isWherein (p)_1x,p_1y) Is the upper edge point P of the hub₁I (x, y) is the gray value of the pixel point (x, y) in the image; correspondingly, lower edge point P₂The coordinate calculation formula isWherein (p)_2x,p_2y) Is the lower edge point P of the hub₂The coordinates of (a);

step b3, passing the upper edge point P of the hub spoke₁And a lower edge point P₂Calculating the spoke center point P of the hub₀Coordinate of (1), hub spoke center point P₀Is calculated by the formulaWherein (p)_0x,p_0y) Is the central point P of the spoke of the hub₀The coordinates of (a);

step b4, passing through the hub central point O and the hub spoke central point P₀Calculating the angle alpha of the spoke of the hub according to the formula

The actual processing process adopts the following steps:

1. the 1 st wheel hub is placed on a rotary table, the rotary table is manually rotated to align the initial polishing point of the wheel hub with the polishing action starting point of the robot, the angle of the wheel hub spoke is calculated through image processing software at the moment and serves as a reference angle for automatic adjustment of the subsequent wheel hub angle, and the reference angle is marked as alpha_r；

2. Manually or automatically feeding the hub from the 2 nd hub, placing the hub on a turntable, calculating the angle of the hub spoke through image processing software, and recording the angle as alpha_f；

3. Calculating the current hub spoke angle alpha_fReference angle alpha with 1 st hub spoke_rA difference of (a), i.e. a ═ a_f-α_r；

4. And sending the wheel hub spoke angle difference value delta alpha to a wheel hub turntable controller, and driving the wheel hub to carry out angle adjustment by the turntable so as to enable the angle of the wheel hub to reach a reference angle.

The invention has the following advantages:

1. the non-contact visual positioning method is adopted to position the hub, so that the positioning precision is improved;

2. through visual positioning, the consistency of the initial positions of the hubs can be ensured, and accurate reference positions are provided for continuous and repeated polishing operation of the robot;

3. the visual positioning algorithm consists of three parts, namely hub central point calculation, hub banner angle calculation and hub angle adjustment;

4. the calculation of the central point of the hub is composed of edge extraction and Hough transformation, so that the precision and the reliability are guaranteed;

5. in the wheel hub spoke angle calculation process, a special upper edge and lower edge extraction algorithm is designed according to the bright spot condition of the wheel hub spoke, and the accuracy and reliability of edge extraction are improved.