A method for detecting vehicle contour information during driving

Abstract

The invention relates to the technical field of vehicle detection, and discloses a method for detecting vehicle contour information during driving of an automobile. During driving of the vehicle, contour information of the vehicle is collected, and a 3D point cloud of a corresponding side is obtained according to matching of the contour information of the vehicle, and a side 3D model of the side of the vehicle contour is constructed according to the 3D point cloud; the collected comparison 3D model is overlapped with an initial 3D model to obtain a non-overlapping area formed by contour changes during driving of the vehicle, and a deviation ratio value of the non-overlapping area of ​​the non-overlapping area is obtained; a contour detection state signal is obtained according to the deviation ratio value of the non-overlapping area, and based on a contour abnormality warning signal, the contour abnormality level of the vehicle during driving is evaluated; based on a contour normal validation signal and in combination with a quality path rate of a current comparison 3D model, a safe driving distance of the vehicle during driving is monitored, and the driving state of the vehicle when the vehicle contour is offset is identified.

Description

Technical Field

[0001] The invention relates to the technical field of vehicle detection, and in particular to a method for detecting vehicle contour information during the driving process of a vehicle. Background Art

[0002] With the development of the automobile industry, there are more and more vehicles, and new requirements for vehicle size and axle number detection have been put forward in transportation.

[0003] For example, patent announcement number CN109238172B discloses a method, device and system for obtaining vehicle contour information, which obtains three-dimensional data information of a detected vehicle in a detection area through a detection device; and obtains contour information of the detected vehicle based on the three-dimensional data information, wherein the contour information includes at least one of the following: length information, width information, and height information.

[0004] For example, patent publication number CN111649678A discloses a dual radar measurement method for vehicle outer contour information, radar calibration, calibrating the length measuring laser radar and the width and height measuring laser radar, initializing the parameters of the length measuring laser radar and the width and height measuring laser radar; performing length measuring target detection and width and height measuring target detection, and calculating the current length, height and width of the vehicle for vehicles that meet the requirements; continuously tracking the target, recording the length, height and width of the vehicle that meets the requirements, and calculating the final length, height and width of the vehicle through continuous multi-frame tracking, which is the vehicle's contour data.

[0005] The above technologies only record the collection and acquisition of vehicle contour information to obtain the actual length, width and height information of the vehicle. However, when the vehicle is driving, the vehicle contour will shift due to driving stability and installation stability. The existing technology lacks the recognition of the vehicle driving state when the vehicle contour shifts, which has great limitations. Summary of the invention

[0006] The purpose of the present invention is to provide a method for detecting vehicle contour information during driving of a vehicle, so as to solve the problem that the driving state of the vehicle cannot be effectively identified when the vehicle contour is offset.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A method for detecting vehicle contour information during driving of a vehicle comprises the following steps:

[0009] During the driving process of the vehicle, the contour information of the vehicle is collected, and the 3D point cloud of the corresponding side is obtained according to the matching of the contour information of the vehicle, and the side 3D model of the side of the vehicle contour is constructed according to the 3D point cloud;

[0010] The side 3D model generated by the first collected vehicle profile information is recorded as the initial 3D model, and the side 3D model obtained by each collected vehicle profile information after the initial 3D model is recorded as the comparison 3D model;

[0011] Overlapping the collected comparison 3D model with the initial 3D model to obtain the non-overlapping area formed by the contour change of the vehicle during driving, and obtaining the deviation ratio value of the non-overlapping area;

[0012] A contour detection status signal is obtained according to the deviation ratio value of the non-overlapping area, and the contour detection status signal includes a contour abnormality warning signal and a contour normal verification signal;

[0013] Based on the contour abnormality warning signal, the contour abnormality level of the vehicle during driving is evaluated;

[0014] Based on the contour normal verification signal and combined with the quality path rate of the current comparison 3D model, the safe driving distance of the vehicle during driving is monitored.

[0015] As a further solution of the present invention: the process of obtaining the bias ratio value of the non-overlapping area is:

[0016] Using the initial 3D model as a base, the collected comparison 3D model is overlapped on the initial 3D model;

[0017] The non-overlapping areas of the comparison 3D model and the initial 3D model are obtained, the volumes of all non-overlapping areas are obtained, and the ratio of the volumes of the non-overlapping areas to the preset volume thresholds of the non-overlapping areas is calculated to obtain the partial ratio values ​​of the non-overlapping areas.

[0018] As a further solution of the present invention: presetting a deviation ratio threshold fi of the non-overlapping area;

[0019] If the deviation ratio Fi of the non-overlapping area is greater than or equal to the deviation ratio threshold fi of the non-overlapping area, the acquisition of the comparison 3D model is recorded as abnormal contour detection, and a contour abnormality warning signal is generated;

[0020] If the deviation ratio Fi of the non-overlapping area is less than the deviation ratio threshold fi of the non-overlapping area, the acquisition of the comparison 3D model is recorded as normal contour detection, and a contour normal verification signal is generated.

[0021] As a further solution of the present invention: based on the contour abnormality warning signal, the area ratio of the non-overlapping area and the fluctuation variance value of the non-overlapping area are obtained;

[0022] The fluctuation variance value of the non-overlapping area is recorded as Fa;

[0023] The area ratio of the non-overlapping area is recorded as Fq;

[0024] Calculate the warning value of the non - overlapping area of the comparison 3D model through the formula .

[0025] As a further solution of the present invention: The limit value of the preset warning value threshold of the non - overlapping area is 1 and 2, where 1 < 2;

[0026] When Aq < Aq1, a first - level warning signal of the vehicle driving state is obtained;

[0027] When Aq1 ≤ Aq < Aq2, a second - level warning signal of the vehicle driving state is obtained

[0028] When Aq ≥ Aq2, a third - level warning signal of the vehicle driving state is obtained.

[0029] As a further solution of the present invention: Obtain the maximum distance value between the comparison 3D model and the initial 3D model in each non - overlapping area, denoted as the distance deviation value;

[0030] Integrate the distance deviation values of all non - overlapping areas to obtain a distance deviation value group;

[0031] Calculate the distance deviation value group according to the variance calculation formula to obtain the fluctuation variance value of the non - overlapping area.

[0032] As a further solution of the present invention: Mark the non - overlapping area between the comparison 3D model and the initial 3D model;

[0033] Project the marked side 3D model perpendicular to the corresponding side to obtain a side 3D projection plane;

[0034] Connect all the marked points on the side 3D projection plane end to end in a clockwise direction to obtain the marked area of the side 3D projection plane;

[0035] Calculate the ratio of the area of the marked area of the side 3D projection plane to the area of the side 3D projection plane to obtain the area ratio of the non - overlapping area.

[0036] As a further solution of the present invention: During the detection period, obtain the road surface quality coefficient of two adjacent monitoring points, denoted as Li;

[0037] Denote the path between two adjacent monitoring points as the monitoring path, and obtain the length of the monitoring path matched by two adjacent monitoring points, denoted as Ci;

[0038] Through the formula ​The quality path rate Lc of the monitoring path is calculated, where i is the number of monitoring paths, i=1, ..., n;

[0039] The quality path rate of the monitoring path is multiplied by the total length of the monitoring path to obtain the effective value of the path quality change.

[0040] As a further solution of the present invention: obtaining the maximum value of the deviation ratio of the non-overlapping areas obtained by processing multiple comparison 3D models within the detection period, and recording it as the effective value of the side trend change within the detection period;

[0041] The effective value of the route mass change is calculated by ratio with the effective value of the side trend change within the detection period to obtain the side change mass-to-speed ratio;

[0042] The difference between the deviation ratio threshold of the non-overlapping area and the effective value of the lateral trend change is calculated to obtain the finite value of the lateral trend change;

[0043] The finite value of the lateral trend change is multiplied by the lateral change mass-to-velocity ratio value to obtain the warning path length value, which is recorded as Ly.

[0044] As a further solution of the present invention: a quality path rate of the early warning path is calculated as a ratio to a quality path rate of the monitoring path to obtain a quality path adjustment ratio;

[0045] If the quality path adjustment ratio is greater than 1, the maximum value of the vehicle's warning path is the product of the warning path length value and the quality path adjustment ratio, recorded as Lymax, and the vehicle's warning path length interval is obtained as [Ly, Lymax];

[0046] If the quality path adjustment ratio is less than or equal to 1, the minimum value of the vehicle's warning path is the product of the warning path length value and the quality path adjustment ratio, denoted as Lymin, and the vehicle's warning path length interval is obtained as [Lymin, Ly].

[0047] Beneficial effects of the present invention:

[0048] The present invention collects the contour information of the vehicle during the driving process, generates a side 3D model according to the collected vehicle information, overlaps the collected comparison 3D model with the initial 3D model, and obtains the non-overlapping area of ​​the comparison 3D model and the initial 3D model. The non-overlapping area refers to the deviation of the cargo on the vehicle compared with the first contour information during the driving process of the vehicle. That is, the larger the deviation ratio of the non-overlapping area is, the greater the deviation of the vehicle contour during the driving process is, and the greater the risk of cargo falling during the driving process is, so as to realize early warning of the vehicle contour information;

[0049] The greater the deviation of the vehicle contour during driving, the greater the contour abnormality warning signal is generated, indicating that there is a very high risk of cargo falling and low driving safety. According to the distance deviation value of the non-overlapping area between the comparison 3D model and the initial 3D model, the greater the distance deviation value of the non-overlapping area, the greater the abnormality of the non-overlapping area. Combined with the overall proportion of the non-overlapping area, the non-overlapping area warning value of the comparison 3D model is calculated, so as to complete the identification of the warning degree of the vehicle driving status according to the non-overlapping area warning value, and realize the visual management of vehicle warnings.

[0050] The present invention processes the road quality coefficients of a plurality of adjacent monitoring points within a detection period and the length of the monitoring path matched by the adjacent monitoring points based on the profile normal verification signal, obtains the quality path rate of the monitoring path, obtains the effective value of the road quality change according to the path length of the monitoring path, obtains the warning path length value based on the effective value of the road quality change and the finite value of the side trend change, calculates the ratio of the quality path rate of the warning path to the quality path rate of the monitoring path, obtains the quality path adjustment ratio, obtains the warning path length interval of the vehicle according to the quality path adjustment ratio, thereby realizing safe operation of the vehicle within the warning path length interval, realizing early warning judgment on vehicle driving safety, and improving driving safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The present invention will be further described below in conjunction with the accompanying drawings.

[0052] Figure 1 This is a flow chart of a method for detecting vehicle contour information during driving of a vehicle according to an embodiment of the present invention;

[0053] Figure 2 This is a flow chart of contour detection state signal recognition in a method for detecting contour information of a vehicle during driving according to an embodiment of the present invention;

[0054] Figure 3 The present invention is a flowchart of a vehicle driving state recognition method in a method for detecting vehicle contour information of a vehicle during driving. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example 1

[0056] See also Figure 1As shown, the present invention is a method for detecting vehicle contour information during driving of a vehicle, comprising the following steps:

[0057] During the driving process of the vehicle, the contour information of the vehicle is collected, and the 3D point cloud of the corresponding side is obtained according to the contour information of the vehicle, and the side 3D model of the side of the vehicle contour is constructed according to the 3D point cloud;

[0058] The side 3D model generated by the first collected vehicle profile information is recorded as the initial 3D model, and the side 3D model obtained by each collected vehicle profile information after the initial 3D model is recorded as the comparison 3D model;

[0059] Overlapping the collected comparison 3D model with the initial 3D model to obtain the non-overlapping area formed by the contour change of the vehicle during driving, and obtaining the deviation ratio value of the non-overlapping area;

[0060] A contour detection status signal is obtained according to the deviation ratio value of the non-overlapping area, and the contour detection status signal includes a contour abnormality warning signal and a contour normal verification signal;

[0061] Based on the contour abnormality warning signal, the contour abnormality level of the vehicle during driving is evaluated;

[0062] Based on the contour normal verification signal and combined with the quality path rate of the current comparison 3D model, the safe driving distance of the vehicle during driving is monitored.

[0063] The vehicle profile information collected for the first time refers to the vehicle profile information collected for the first time by the vehicle profile information collection device when the vehicle enters the highway;

[0064] The vehicle profile information collected each time after the initial 3D model refers to the vehicle profile information collected by each subsequent vehicle profile information collection device when the vehicle is on the highway;

[0065] The vehicle profile information includes vehicle length information, vehicle width information, and vehicle height information;

[0066] Among them, the side 3D model is mainly the side of the vehicle during driving.

[0067] The process of obtaining the bias ratio value of the non-overlapping area is as follows:

[0068] The obtained comparison 3D models of the same side are compared with the initial 3D model respectively. Specifically:

[0069] Using the initial 3D model as a base, the collected comparison 3D model is overlapped on the initial 3D model;

[0070] Obtaining non-overlapping areas of the comparison 3D model and the initial 3D model, obtaining the volumes of all non-overlapping areas, calculating the ratio of the volumes of the non-overlapping areas to a preset volume threshold of the non-overlapping areas, and obtaining the partial ratio values ​​of the non-overlapping areas;

[0071] The preset volume threshold of the non-overlapping area is an empirical value, which is set by the staff according to the size of the side of the vehicle outline. The larger the size of the side of the vehicle outline, the larger the corresponding preset volume threshold of the non-overlapping area.

[0072] Obtain the deviation ratio value Fi of the non-overlapping area in each comparison of the 3D model;

[0073] Compare the deviation ratio value Fi of the non-overlapping area of ​​the side with the deviation ratio threshold value fi of the corresponding non-overlapping area;

[0074] See also Figure 2 If the deviation ratio Fi of the non-overlapping area is greater than or equal to the deviation ratio threshold fi of the non-overlapping area, the acquisition of the comparison 3D model is recorded as abnormal contour detection, and a contour abnormality warning signal is generated;

[0075] If the deviation ratio Fi of the non-overlapping area is less than the deviation ratio threshold fi of the non-overlapping area, the acquisition of the comparison 3D model is recorded as normal contour detection, and a contour normal verification signal is generated.

[0076] The technical solution of the present invention is as follows: during the driving process of the vehicle, the contour information of the vehicle is collected, and a side 3D model is generated based on the collected vehicle information, and the collected comparison 3D model is overlapped with the initial 3D model to obtain a non-overlapping area between the comparison 3D model and the initial 3D model, and the non-overlapping area refers to the deviation of the cargo on the vehicle compared to the first contour information during the driving process of the vehicle, that is, the larger the deviation ratio of the non-overlapping area, the greater the deviation of the vehicle contour during the driving process. Example 2

[0077] On the basis of Example 1, based on the contour abnormality warning signal, the vehicle contour warning degree during vehicle driving is evaluated, specifically:

[0078] Mark the non-overlapping areas of the comparison 3D model and the initial 3D model, and obtain the positions of the marking points on the side 3D model;

[0079] Obtain the maximum distance between the comparison 3D model and the initial 3D model in each non-overlapping area, recorded as the distance deviation value;

[0080] Integrate the distance deviation values ​​of all non-overlapping areas to obtain a distance deviation value group;

[0081] Calculate the distance deviation value group according to the variance calculation formula to obtain the fluctuation variance value of the non-coincident area, denoted as Fa;

[0082] Project the marked side 3D model perpendicular to the corresponding side to obtain the side 3D projection plane;

[0083] Connect all the marked points on the side 3D projection plane end to end in a clockwise direction to obtain the marked area of the side 3D projection plane;

[0084] Obtain the area of the marked area of the side 3D projection plane and the area of the side 3D projection plane, and calculate the ratio of the area of the marked area of the side 3D projection plane to the area of the side 3D projection plane to obtain the area ratio of the non-coincident area, denoted as Fq;

[0085] Through the formula Calculate the warning value of the non-coincident area of the comparison 3D model ;

[0086] The limit values of the warning value threshold of the non-coincident area are preset as 1 and 2, where 1 < 2;

[0087] Among them, the limit values 1 and 2 of the warning value threshold of the non-coincident area are empirical values, obtained according to experience:

[0088] In the actual process of obtaining, there are many groups of area ratios of non-coincident areas and fluctuation variance values of non-coincident areas. Process many groups of area ratios of non-coincident areas and fluctuation variance values of non-coincident areas to obtain the corresponding warning values of non-coincident areas. The staff identify the vehicle driving state level based on so many groups of warning values of non-coincident areas, so as to obtain a corresponding relationship between the warning value of the non-coincident area and the vehicle driving state level. Furthermore, deduce and divide the warning value threshold of the non-coincident area according to the vehicle driving state level of the warning value of the non-coincident area, so as to obtain the limit values 1 and 2 of the warning value threshold of the non-coincident area. By comparing the limit values of the warning value threshold of the non-coincident area, the identification of the vehicle driving state level corresponding to the warning value of the non-coincident area is completed;

[0089] Refer to Figure 3 , when Aq < Aq1, obtain the first-level warning signal of the vehicle driving state;

[0090] When Aq1 ≤ Aq < Aq2, obtain the second-level warning signal of the vehicle driving state

[0091] When Aq≥Aq2, a third-level warning signal of the vehicle driving status is obtained;

[0092] Among them, the vehicle warning level corresponding to the third-level warning signal of the vehicle driving status is higher than the vehicle warning level corresponding to the second-level warning signal, and the vehicle warning level corresponding to the second-level warning signal of the vehicle driving status is higher than the vehicle warning level corresponding to the first-level warning signal. The higher the vehicle warning level, the more dangerous the vehicle driving is, and the more likely the goods will be scattered or dropped, so as to realize the visual management of vehicle warnings;

[0093] The technical solution of the present invention is as follows: when the deviation of the vehicle contour during driving is greater, so that a contour abnormality warning signal is generated, it means that there is a very high risk of cargo falling and the driving safety is low. According to the distance deviation value of the non-overlapping area between the comparison 3D model and the initial 3D model, the larger the distance deviation value of the non-overlapping area, the larger the abnormality of the non-overlapping area. Combined with the overall proportion of the non-overlapping area, the non-overlapping area warning value of the comparison 3D model is calculated, thereby completing the identification of the warning degree of the vehicle driving status according to the non-overlapping area warning value. Example 3

[0094] On the basis of Example 1, based on the contour normal verification signal, the safe distance of the vehicle is monitored to achieve vehicle online warning;

[0095] When a vehicle is driving, especially on a highway, there are usually multiple monitoring points for the vehicle profile information. Each monitoring point corresponds to a comparative 3D model that matches the license plate profile information.

[0096] A detection period is preset, and the maximum value of the deviation ratio of non-overlapping areas obtained by processing multiple comparison 3D models within the detection period is obtained, which is recorded as the effective value of the side trend change within the detection period and recorded as Aq;

[0097] During the detection period, the road quality coefficients of two adjacent monitoring points are obtained, and the path between the two adjacent monitoring points is recorded as the monitoring path;

[0098] That is, by obtaining the road surface flatness, damage condition, load-bearing capacity and anti-skid capacity indicators of the monitoring path;

[0099] The road surface flatness of the monitoring path is marked as P1;

[0100] The damage condition of the monitoring path is marked as P2;

[0101] The carrying capacity of the monitoring path is marked as P3;

[0102] The anti-slip capacity index of the monitoring path is marked as P4;

[0103] Weighted processing is performed on the road surface smoothness, damage condition, load-bearing capacity and anti-skid capacity indicators of the monitoring path;

[0104] That is, the pavement quality coefficient Li is obtained by the formula Li=P1Xd1+P2Xd2+P3Xd3+P4Xd4, i is the number of monitoring paths, i=1, ..., n, where d1, d2, d3, d4 represent the weight values ​​of the pavement flatness, damage condition, bearing capacity and anti-skid ability indicators, respectively, d1+d2+d3+d4=1;

[0105] Get the length of the monitoring path matched by two adjacent monitoring points, denoted as Ci, where i is the number of monitoring paths, i=1, ..., n;

[0106] By formula The quality path rate Lc of the monitoring path is calculated;

[0107] The quality path rate of the monitoring path is multiplied by the total length of the monitoring path to obtain the effective value of the path quality change;

[0108] Among them, the total monitoring path length is the sum of the path lengths of multiple monitoring paths within the detection period;

[0109] The effective value of the route mass change is calculated by ratio with the effective value of the side trend change within the detection period to obtain the side change mass-to-speed ratio;

[0110] The difference between the deviation ratio threshold of the non-overlapping area and the effective value of the lateral trend change is calculated to obtain the finite value of the lateral trend change;

[0111] The finite value of the side trend change is multiplied by the mass-velocity ratio of the side change to obtain the warning path length value, which is recorded as Ly;

[0112] According to the warning path length value, the road quality coefficient of the vehicle traveling in the warning path is obtained, and the quality path rate of the warning path is obtained in the same manner as the quality path rate of the monitoring path;

[0113] The quality path rate of the early warning path is calculated by ratio with the quality path rate of the monitoring path to obtain the quality path adjustment ratio;

[0114] Identify the quality path adjustment ratio:

[0115] If the quality path adjustment ratio is greater than 1, the maximum value of the vehicle's warning path is the product of the warning path length value and the quality path adjustment ratio, recorded as Lymax, and the vehicle's warning path length interval is obtained as [Ly, Lymax];

[0116] If the quality path adjustment ratio is less than or equal to 1, the minimum value of the vehicle's warning path is the product of the warning path length value and the quality path adjustment ratio, denoted as Lymin, and the vehicle's warning path length interval is obtained as [Lymin, Ly];

[0117] This enables the vehicle's driving distance to be monitored based on the vehicle's warning path length interval, ensuring vehicle driving safety and preventing cargo from being scattered due to changes in the vehicle's contour during driving, thereby avoiding unnecessary driving risks.

[0118] The technical solution of the present invention is as follows: the road quality coefficients of multiple adjacent monitoring points within a detection period are processed with the length of the monitoring path matched by the adjacent monitoring points to obtain the quality path rate of the monitoring path, and then the effective value of the road quality change is obtained according to the path length of the monitoring path, and the warning path length value is obtained based on the effective value of the road quality change and the finite value of the side trend change, and the quality path rate of the warning path is calculated to be a ratio with the quality path rate of the monitoring path to obtain the quality path adjustment ratio, and the warning path length interval of the vehicle is obtained according to the quality path adjustment ratio, so as to realize the safe operation of the vehicle within the warning path length interval and realize early warning judgment on the vehicle driving safety.

[0119] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for detecting vehicle contour information during driving, characterized in that: It includes the following steps: During the vehicle driving process, collect the contour information of the vehicle, match the corresponding side 3D point cloud according to the contour information of the vehicle, and construct a side 3D model of the vehicle contour side based on the 3D point cloud; Record the side 3D model generated from the first collected vehicle contour information as the initial 3D model, and record the side 3D model obtained from the vehicle contour information collected each time after the initial 3D model as the comparison 3D model; Overlap the collected comparison 3D model with the initial 3D model to obtain the non-overlapping area formed by the contour change of the vehicle contour information during driving, and obtain the deviation ratio of the non-overlapping area; Obtain the contour detection status signal according to the deviation ratio of the non-overlapping area. The contour detection status signal includes a contour abnormality warning signal and a contour normal verification signal; Based on the contour abnormality warning signal, evaluate the contour abnormality level of the vehicle during driving; Based on the contour normal verification signal, combined with the quality path rate of the current comparison 3D model, monitor the safe driving distance of the vehicle during driving; During the detection period, obtain the road surface quality coefficients of two adjacent monitoring points, denoted as Li; Denote the path between two adjacent monitoring points as the monitoring path, and obtain the length of the monitoring path matched by two adjacent monitoring points, denoted as Ci; By formula The quality path rate Lc of the monitoring path is calculated, where i is the number of monitoring paths, i=1, ..., n; Perform a product calculation on the quality path rate of the monitoring path and the total length of the monitoring path to obtain the effective value of the road surface quality change.

2. The method for detecting vehicle contour information during driving of a vehicle according to claim 1, characterized in that: The process of obtaining the deviation ratio of the non-overlapping area is as follows: Taking the initial 3D model as the base, overlap the collected comparison 3D model on the initial 3D model; Obtain the non-overlapping area between the comparison 3D model and the initial 3D model, obtain the volume of all non-overlapping areas, and perform a ratio calculation on the volume of the non-overlapping area and the preset volume threshold of the non-overlapping area to obtain the deviation ratio of the non-overlapping area.

3. The method for detecting vehicle contour information during driving of a vehicle according to claim 1, characterized in that: Preset the deviation ratio threshold fi of the non-overlapping area; If the deviation ratio Fi of the non-overlapping area is greater than or equal to the deviation ratio threshold fi of the non-overlapping area, record the collection of the comparison 3D model this time as an abnormal contour detection and generate a contour abnormality warning signal; If the deviation ratio Fi of the non-overlapping area is less than the deviation ratio threshold fi of the non-overlapping area, record the collection of the comparison 3D model this time as a normal contour detection and generate a contour normal verification signal.

4. The method for detecting vehicle contour information during driving of a vehicle according to claim 1, characterized in that: Based on the contour abnormality warning signal, obtain the area ratio of the non-overlapping area and the fluctuation variance value of the non-overlapping area; Denote the fluctuation variance value of the non-overlapping area as Fa; Denote the area ratio of the non-overlapping area as Fq; Calculate the non-overlapping area warning value Aq of the comparison 3D model through the formula Aq = Fa * Fq.

5. The method for detecting vehicle contour information during driving of a vehicle according to claim 4, characterized in that: Preset the limit values of the non-overlapping area warning value threshold as Aq1 and Aq2, where Aq1 < Aq2; When Aq < Aq1, obtain a first-level warning signal of the vehicle driving state; When Aq1 ≤ Aq < Aq2, obtain a second-level warning signal of the vehicle driving state; When Aq ≥ Aq2, obtain a third-level warning signal of the vehicle driving state.

6. The method for detecting vehicle contour information during driving of a vehicle according to claim 4, characterized in that: Obtain the maximum spacing value between the comparison 3D model and the initial 3D model in each non-overlapping area, denoted as the distance deviation value; Integrate the distance deviation values ​​of all non-overlapping areas to obtain a distance deviation value group; The distance deviation value group is calculated according to the variance calculation formula to obtain the fluctuation variance value of the non-overlapping area.

7. The method for detecting vehicle contour information during driving of a vehicle according to claim 4, characterized in that: Mark the non-overlapping areas of the comparison 3D model and the initial 3D model; Project the marked side 3D model perpendicularly to the corresponding side to obtain the side 3D projection surface; All the marking points on the side 3D projection surface are connected clockwise from beginning to end to obtain the marking area of ​​the side 3D projection surface; The area ratio of the non-overlapping area is calculated by calculating the ratio of the area of ​​the marked area of ​​the side 3D projection surface to the area of ​​the side 3D projection surface.

8. The method for detecting vehicle contour information during driving of a vehicle according to claim 1, characterized in that: Obtain the maximum value of the deviation ratio of the non-overlapping area obtained by processing multiple comparison 3D models within the detection period, and record it as the effective value of the side trend change within the detection period; The effective value of the route mass change is calculated by ratio with the effective value of the side trend change within the detection period to obtain the side change mass-to-speed ratio; The difference between the deviation ratio threshold of the non-overlapping area and the effective value of the lateral trend change is calculated to obtain the finite value of the lateral trend change; The finite value of the lateral trend change is multiplied by the lateral change mass-to-velocity ratio value to obtain the warning path length value, which is recorded as Ly.

9. The method for detecting vehicle contour information during driving of a vehicle according to claim 8, characterized in that: The quality path rate of the early warning path is calculated by ratio with the quality path rate of the monitoring path to obtain the quality path adjustment ratio; If the quality path adjustment ratio is greater than 1, the maximum value of the vehicle's warning path is the product of the warning path length value and the quality path adjustment ratio, recorded as Lymax, and the vehicle's warning path length interval is obtained as [Ly, Lymax]; If the quality path adjustment ratio is less than or equal to 1, the minimum value of the vehicle's warning path is the product of the warning path length value and the quality path adjustment ratio, denoted as Lymin, and the vehicle's warning path length interval is obtained as [Lymin, Ly).

Images