A truck suspension axle lifting state detection method based on image processing technology

Abstract

The application discloses a truck suspension axle lifting state detection method based on image processing technology, mainly containing the following steps: first, acquiring a truck side image, detecting a tire by using a deep learning method, and acquiring an axle axis and an axle number; for a truck with two axles or more, sorting the axle axes according to the horizontal coordinates according to the position of the truck head; determining the suspension state of each intermediate axle according to the angle relationship of the adjacent axle axes starting from the truck head, and then determining the suspension state of the last axle of the truck; finally, counting the number of axles in the suspension state and the axle serial number, and giving the axle suspension state information. The application can automatically detect the suspension axle state of the truck without manual intervention, can effectively identify whether the truck has a suspension axle in the lifting state and can instantly count, effectively solves the deficiency that the traditional technology is difficult to detect the suspension axle, and thus prevents the malicious fee evasion behavior of the truck driver at the toll station caused by the lifting of the suspension axle.

Description

Technical Field

[0001] The present invention relates to the field of image processing, and in particular to a method for detecting the lifting state of a truck suspension axle based on image processing technology. Background Art

[0002] The suspended axle is a structure commonly found on trucks. When the truck is empty or lightly loaded, the suspended axle can be raised to reduce tire wear and fuel consumption.

[0003] According to the industry standard "Vehicle Type Classification for Toll Road Vehicles" (JT / T 489-2019), truck types are classified based on factors such as the total number of axles (including suspended axles), vehicle length, and gross vehicle mass. To evade tolls, some drivers deliberately lift suspended axles when weighing at toll stations, artificially "losing" them and paying less. However, current detection technology at traditional toll stations has limitations in combating this type of toll evasion. Summary of the Invention

[0004] In view of the shortcomings of the background technology, the present invention provides a method for detecting the lifting state of a truck suspension axle based on image processing technology, which includes the following steps:

[0005] Step S1, obtaining a side image of a truck that can display all tires on one side of the truck;

[0006] Step S2: Detect tires from the truck side image using a deep learning method to obtain the axle center and the number of axles;

[0007] Step S3: If the number of axles is greater than 2, proceed to the next step; otherwise, there is no suspended axle, and go to step S10;

[0008] Step S4, obtaining the position information of the front of the truck in the side image of the truck;

[0009] Step S5: If the truck head is located on the left side of the image, the axle centers are arranged in ascending order according to the horizontal coordinates; otherwise, the axle centers are arranged in descending order according to the horizontal coordinates;

[0010] Step S6, starting from the first axle of the truck's front, sequentially connect the axis centers of two adjacent axles and calculate the angle of each connecting line;

[0011] Step S7, determining the suspension state of each intermediate shaft starting from the front of the truck based on the angular relationship between the connecting lines of adjacent shaft centers;

[0012] Step S8: If the number of non-suspended axles is greater than or equal to 2, proceed to the next step; otherwise, the last axle of the truck, i.e., the end axle, is determined to be a non-suspended axle, and the process goes to step S10;

[0013] Step S9, determining the suspension state of the last axle according to the relationship between the angle of the axis center line of the last two non-suspension axles and the angle of the axis center line of the last non-suspension axle and the last axle;

[0014] Step S10, counting the number of axles in the suspension state and the corresponding axle number;

[0015] Step S11, giving the axle suspension state information;

[0016] The intermediate axle in the step S7 is the axle other than the first axle and the last axle, and the specific method for determining the suspension state of each intermediate axle from the direction of the truck head is as follows: the axis center of the current axle to be determined is P2, the axis center of the previous axle adjacent to the current axle is P1, the axis center of the next axle adjacent to the current axle is P3, the angle of the line connecting P1 and P2 is a1, and the angle of the line connecting P2 and P3 is a2; when the truck head is located on the left side of the image, if a1

[0017] Compared with the background art, the present application has the beneficial effects that: the present application detects the lifting state of the suspension axle of the truck based on the image processing technology, and can automatically detect the suspension axle state of the truck without manual intervention. Through the deep learning technology for detecting the tire and identifying the number and position of the axle, the present application can effectively identify whether the truck has the suspension axle in the lifting state, and can instantly count, thereby effectively solving the deficiency of the traditional toll station technology that cannot detect the suspension axle, and preventing the escape and evasion of the toll caused by the lifting of the suspension axle. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is the step flow chart of the method of the present application;

[0019] Figure 2 is the truck side image of the embodiment;

[0020] Figure 3 is the tire detection result and axle axis center marking image of the embodiment;

[0021] Figure 4 is the axis center line angle calculation schematic diagram;

[0022] Figure 5 is the intermediate axle suspension state determination schematic diagram;

[0023] Figure 6is an image of the axle suspension state detection result of the embodiment;

[0024] Figure 7 This is the image of the axle suspension state detection result of Example 2;

[0025] Figure 8 This is the axle suspension state detection result image of Example 3;

[0026] Figure 9 This is the image of the axle suspension state detection result of Example 4;

[0027] Figure 10 This is the image of the axle suspension state detection result of Example 5. DETAILED DESCRIPTION

[0028] The present invention will be further described below with reference to the accompanying drawings.

[0029] like Figure 1 As shown, the steps of the method of the present invention are:

[0030] Step S1: Obtain a side view of the truck that shows all tires on one side of the truck

[0031] Get the side image of the truck, such as Figure 2 As shown, this image shows all the tires on one side of the truck.

[0032] Step S2: Use deep learning methods to detect tires from the side image of the truck and obtain the axle center and axle number

[0033] Using a trained deep learning tire detection model, we detect tires from a truck's side profile image. The center of the tire's circumscribed rectangle is determined as the axle center. The axle center coordinates are the image coordinates of the truck's side profile image, with the origin located in the upper left corner of the image, positive horizontal coordinates pointing rightward, and positive vertical coordinates pointing downward. Upon tire detection, the number of axles on the truck can be determined.

[0034] The tire circumscribed rectangle and the axis marked with “×” detected in this embodiment are as follows: Figure 3 As shown, the truck has 6 axles.

[0035] Step S3: If the number of axles is greater than 2, proceed to the next step; otherwise, there is no suspended axle, go to step S10

[0036] According to the number of axles detected in step S2, if the number of axles is greater than 2, there may be a suspended axle, and the next step is performed; otherwise, there is no suspended axle, and the process goes to step S10.

[0037] In this embodiment, the number of axles detected is 6, and there may be a suspended axle, so it is necessary to proceed to step S4 for subsequent processing.

[0038] Step S4, obtaining the position information of the truck head in the truck side image

[0039] The position information of the truck head in the truck side image is obtained. The position of the truck head in the truck side image is the position of the truck head relative to the truck tail in the truck side image. If the truck head is on the left side of the truck tail, the truck head is on the left side of the image; if the truck head is on the right side of the truck tail, the truck head is on the right side of the image.

[0040] In this embodiment, the truck head is on the left side of the truck tail, so it is determined that the truck head is on the left side of the image.

[0041] Step S5, if the truck head is on the left side of the image, the horizontal coordinates of the axle shafts are arranged in ascending order; otherwise, the horizontal coordinates of the axle shafts are arranged in descending order

[0042] According to the position of the truck head obtained in step S4 and the coordinates of the axle shafts obtained in step S2, the horizontal coordinates of the axle shafts are sorted. If the truck head is on the left side of the image, the horizontal coordinates of the axle shafts are arranged in ascending order; otherwise, the horizontal coordinates of the axle shafts are arranged in descending order.

[0043] In this embodiment, since the truck head is on the left side of the image, the horizontal coordinates of the axle shafts need to be arranged in ascending order.

[0044] Step S6, starting from the first axle of the truck head, the shaft centers of adjacent two axles are connected in turn and the angles of the connecting lines are calculated

[0045] According to the axle shafts obtained in step S2, starting from the first axle of the truck head, the shaft centers of adjacent two axles are connected in turn and the angles of the connecting lines are calculated. The angle of the connecting line of the shaft centers of adjacent two axles is calculated by the following method: rotate the horizontal line clockwise or counterclockwise by an acute angle around any one of the shaft centers of the two axles, so that the horizontal line coincides with the connecting line or its extension of the shaft centers of the two axles, and the angle of rotation is the angle of the connecting line of the shaft centers of the two axles, taking positive for clockwise rotation and negative for counterclockwise rotation, as shown in Figure 4 .

[0046] Step S7, determining the suspension state of each axle of the intermediate axles from the direction of the truck head according to the angle relationship of the adjacent shaft centers

[0047] The other axles except the first axle and the last axle (i.e. the last axle) are intermediate axles. The specific method for determining the suspension state of each intermediate axle starting from the front direction of the truck is as follows: let the axis center of the current axle to be determined to be in suspension state be P2, the axis center of the previous axle adjacent to the current axle be P1, the axis center of the next axle adjacent to the current axle be P3, the angle of the line connecting P1 and P2 be α1, and the angle of the line connecting P2 and P3 be α2; when the front of the truck is on the left side of the image, if α1<α2 and |α1-α2|>θ, then the current axle is determined to be in suspension state, and this axle is a suspended axle; otherwise, the current axle is determined to be in non-suspension state, and this axle is a non-suspension axle; when the front of the truck is on the right side of the image, if α1>α2 and |α1-α2|>θ, then the current axle is determined to be in suspension state, and this axle is a suspended axle; otherwise, the current axle is determined to be in non-suspension state, and this axle is a non-suspension axle; where θ is the set angle threshold, such as Figure 5 shown.

[0048] In this embodiment, the truck is a 6-axle truck, the front of the truck is located on the left side of the image, and the angle threshold θ is defined as 7°. The angle α1 of the line connecting the axis center of the first axis and the axis center of the second axis is -15.6°, the angle α2 of the line connecting the axis center of the second axis and the axis center of the third axis is 1.7°, |α1-α2| is 17.3°>θ, and α1<α2, so the second axis is determined to be a suspended axis; the angle α3 of the line connecting the axis center of the third axis and the axis center of the fourth axis is -11.1°, |α2–α3| is 12.8°>θ but α2>α3, so the third axis is determined to be a non-suspended axis; the angle α4 of the line connecting the axis center of the fourth axis and the axis center of the fifth axis is -10.6°, |α3–α4| is 0.5°<θ, so the fourth axis is determined to be a non-suspended axis; the angle α5 of the line connecting the axis center of the fifth axis and the axis center of the sixth axis is -8.6°, |α4–α5| is 2.0°<θ, so the fifth axis is determined to be a non-suspended axis.

[0049] Step S8: If the number of non-suspended axles is greater than or equal to 2, proceed to the next step; otherwise, the last axle of the truck is determined to be a non-suspended axle, and the process goes to step S10.

[0050] If the number of axles other than the final axle that are determined to be non-suspended axles is greater than or equal to two, the final axle may be either a non-suspended axle or a suspended axle in a lifted state, and further determination is required in step S9. Otherwise, the final axle is determined to be a non-suspended axle, and the process proceeds directly to step S10.

[0051] In this embodiment, the number of axles other than the final axle that are determined to be non-suspended axles is 4, including 3 intermediate axles and 1 first axle (the first axle starting from the front of the vehicle cannot be a suspended axle and can only be a non-suspended axle). The condition that the number of non-suspended axles is greater than or equal to 2 is met, so it is necessary to proceed to step S9 to determine the status of the final axle.

[0052] Step S9: Determine the suspension state of the last axis based on the relationship between the angle between the axis centers of the last two non-suspended axes and the angle between the axis centers of the last non-suspended axis and the last axis.

[0053] The specific method for determining the suspension state of the final axis is as follows: let the axis center of the penultimate non-suspended axis be C1, the axis center of the second-to-last non-suspended axis be C2, and the axis center of the final axis be C3. The angle between C1 and C2 is β1, and the angle between C2 and C3 is β2. If |β1-β2|>θ, then the final axis is determined to be in a suspension state; otherwise, the final axis is determined to be in a non-suspended state; where θ is a set angle threshold, which is the same as the angle threshold set in step S7.

[0054] In this embodiment, the angle β1 between the axis centers of the penultimate non-suspended axis and the second-to-last non-suspended axis is -10.4°, and the angle β2 between the axis centers of the penultimate non-suspended axis and the last axis is -8.6°. Since |β1-β2|=1.8°<θ, the last axis, i.e., the sixth axis, is a non-suspended axis.

[0055] The axle suspension state detection results of this embodiment are as follows Figure 6 As shown, the axis with the axis center marked as "O" plus "+" is a suspended axis in a lifted state.

[0056] Step S10: Count the number of axles in the suspended state and the corresponding axle numbers

[0057] Based on the axle suspension state detection results from the previous steps, the number of axles in the suspension state is counted, and the corresponding axle numbers in the suspension state are counted. The axle numbers start with the first axle as 1 and increase by 1 from the first axle to the last axle.

[0058] In this embodiment, the truck has one suspended axle in a lifted state, and the corresponding axle number is 2.

[0059] Step S11: Provide axle suspension status information

[0060] Finally, the suspension status information of the axles of the truck under test is given, including the number of axles, the number of suspended axles in the lifted state, and the axle serial number.

[0061] In this embodiment, the axle suspension state information is provided as follows: the number of axles is 6, the number of suspended axles in the lifting state is 1, and the sequence number is 2.

[0062] In addition to the above-mentioned embodiments, the present invention also provides Figures 7 to 10 The remaining 4 embodiments are shown. Figure 7 The figure shows the axle suspension state detection result of Example 2. This example is a 4-axle truck with the front end located on the left, and the first intermediate axle and the end axle are suspended axles in a raised state. Figure 8The figure shows the axle suspension state detection result of Example 3. This example is a 4-axle truck with the front end located on the left, and the rear axle is a suspended axle in a raised state. Figure 9 The figure shows the axle suspension state detection result of Example 4, which is a 5-axle truck with the front of the truck located on the right side and no suspended axle in the raised state. Figure 10 The figure shows the axle suspension state detection result of Example 6. This example is a three-axle truck with the front of the truck located on the right side, and the middle axle is a suspended axle in a raised state.

[0063] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified, or some technical features can be replaced by equivalents without departing from the spirit of the technical solutions of the present invention, which should all be included in the scope of the technical solutions claimed for protection by the present invention.

Claims

1. A method for detecting the lifting state of a truck suspension axle based on image processing technology, characterized by: The following steps are involved: S1: Acquire a side image of a truck showing all tires on one side of the truck; S2: Use deep learning methods to detect tires from truck side images and obtain axle centers and the number of axles; S3: If the number of axles is greater than 2, proceed to the next step; otherwise, there is no suspended axle, go to S10; S4: Obtaining the position information of the truck head in the truck side image; S5: If the truck head is on the left side of the image, the axle centers are sorted in ascending order according to the horizontal coordinates; otherwise, the axle centers are sorted in descending order according to the horizontal coordinates; S6: Starting from the first axle at the front of the truck, connect the axis centers of two adjacent axles in sequence and calculate the angle of each connecting line; S7: Determine the suspension state of each intermediate shaft starting from the front of the truck based on the angular relationship between the adjacent shaft centers; S8: If the number of non-suspended axles is greater than or equal to 2, proceed to the next step; otherwise, determine that the last axle of the truck, i.e., the end axle, is a non-suspended axle, and go to S10; S9: Determine the suspension state of the last axis according to the relationship between the angle between the axis centers of the last two non-suspended axes and the angle between the axis centers of the last non-suspended axis and the last axis; S10: Count the number of axles in the suspended state and the corresponding axle numbers; S11: Provide axle suspension status information; The intermediate axes in step S7 are axles other than the first and last axes. The specific method for determining the suspension state of each intermediate axis starting from the front of the truck is as follows: let the axis center of the current axis to be determined as suspension state be P2, the axis center of the previous axis adjacent to the current axis be P1, the axis center of the next axis adjacent to the current axis be P3, the angle of the line connecting P1 and P2 be α1, and the angle of the line connecting P2 and P3 be α2; when the front of the truck is located on the left side of the image, if α1<α2 and |α1-α2|>θ, then the current axis is determined to be in suspension state and this axis is a suspended axis; otherwise, the current axis is determined to be in non-suspension state and this axis is a non-suspension axis; when the front of the truck is located on the right side of the image, if α1>α2 and |α1-α2|>θ, then the current axis is determined to be in suspension state and this axis is a suspended axis; otherwise, the current axis is determined to be non-suspension state and this axis is a non-suspension axis; where θ is a set angle threshold.

2. The method for detecting the lifting state of a truck suspension axle based on image processing technology according to claim 1, characterized in that: In step S2, the axle center is determined by using a deep learning method to obtain the center of the circumscribed rectangle of the tire detected from the side image of the truck. The axis center coordinates are the image coordinates of the side image of the truck, that is, the origin is located in the upper left corner of the image, the horizontal coordinate is positive to the right, and the vertical coordinate is positive downward.

3. The method for detecting the lifting state of a truck suspension axle based on image processing technology according to claim 1, characterized in that: In step S4, the position of the front of the truck in the side image of the truck is the position of the front of the truck relative to the rear of the truck in the side image of the truck. If the front of the truck is on the left side of the rear of the truck, the front of the truck is located on the left side of the image; if the front of the truck is on the right side of the rear of the truck, the front of the truck is located on the right side of the image.

4. The method for detecting the lifting state of a truck suspension axle based on image processing technology according to claim 1, characterized in that: In step S6, the angle of the line connecting the centers of the two adjacent axles is calculated by the following method: a horizontal line is rotated clockwise or counterclockwise around either of the two axle centers by an acute angle so that the horizontal line coincides with the line connecting the centers of the two axles or its extension. The angle of rotation is the angle of the line connecting the centers of the two adjacent axles, which is positive for clockwise rotation and negative for counterclockwise rotation.

5. The method for detecting the lifting state of a truck suspension axle based on image processing technology according to claim 1, characterized in that: The specific method for determining the suspension state of the final axis in step S9 is as follows: assuming that the axis center of the penultimate non-suspended axis is C1, the axis center of the second-to-last non-suspended axis is C2, and the axis center of the final axis is C3, the angle of the line connecting C1 and C2 is β1, and the angle of the line connecting C2 and C3 is β2. If |β1-β2|>θ, the final axis is determined to be in a suspension state; otherwise, the final axis is determined to be in a non-suspended state; where θ is a set angle threshold, which is the same as the angle threshold set in step S7.

Images