CN114088041B - Vehicle three-dimensional scanning imaging method and system - Google Patents

Abstract

The invention belongs to the technical field of image recognition, and discloses a vehicle three-dimensional scanning imaging method, which comprises the following steps: establishing a unified three-dimensional coordinate system according to the operation environment of the vehicle; the imaging equipment is used for carrying out motion scanning on the vehicles parked in the parking area, so as to obtain the appearance image data of the vehicles at different moments and uniformly present the appearance image data in the three-dimensional coordinate system; and aggregating the outline image data presented in the unified coordinate system to form complete three-dimensional image data of the vehicle outline. The invention solves the problems of low three-dimensional imaging precision, poor imaging quality and high cost of the original vehicle.

Description

Vehicle three-dimensional scanning imaging method and system

Technical Field

The present invention belongs to the technical field of image recognition, and in particular relates to a vehicle three-dimensional scanning imaging method and system.

Background Art

With the development of urban modernization, automatic car washing services have been widely demanded by society.

In the field of automatic car washing, it is necessary to identify and image the appearance of the vehicle in order to form subsequent path planning. Most of the current three-dimensional imaging methods have problems such as poor imaging data quality and high cost. For example: imaging with a beam tube: only two-dimensional data, and the resolution is limited by the principle of the beam tube and cannot be too high, and the cost is high; fixed-position imaging: single or multiple ranging or imaging sensors (ranging sensors, ordinary cameras, three-dimensional cameras, lidar), the sensor incident angle and range are not ideal, the cost is extremely high, and the data accuracy and resolution of ranging sensors and ordinary cameras are poor; space roaming imaging: SLAM-style imaging, low accuracy, poor data consistency, poor incidence angle and range. Poor incidence angle and poor range can easily lead to poor imaging accuracy and poor imaging quality.

In addition, the vehicle appearance data obtained by imaging sensors often have missing and abnormal problems, which will also lead to poor imaging effects, especially in auxiliary parts such as dark mirrors, car windows and antennas, where the imaging quality is prone to defects.

Therefore, how to design a vehicle three-dimensional scanning imaging method and system with high imaging accuracy, good imaging quality and low cost is a technical problem that needs to be solved at present.

Summary of the invention

The purpose of the present invention is to provide a vehicle three-dimensional scanning imaging method to solve the problems of low recognition accuracy, poor imaging quality and high cost of the original vehicle three-dimensional imaging.

To this end, a first aspect of the present invention provides a vehicle three-dimensional scanning imaging method, the method comprising:

Establish a unified three-dimensional coordinate system based on the working environment of the vehicle;

Performing motion scanning on vehicles parked in the parking area by using an imaging device to obtain image data of the vehicle's appearance at different times and presenting them uniformly in the unified three-dimensional coordinate system;

The appearance image data presented in the three-dimensional coordinate system are aggregated to form complete three-dimensional image data of the vehicle.

Furthermore, the method further comprises: judging the point cloud quality of the scanned shape image data and performing optimization processing on those with quality defects, wherein the optimization processing comprises:

Determine the continuity of the vehicle appearance image data and perform fuzzy filling processing on the blank image areas; and/or,

According to the symmetry characteristics of the vehicle, when it is detected that the image data obtained by scanning one side of the vehicle is poor, the qualified image data obtained by scanning the other side is mirrored, and finally the complete three-dimensional image data of the vehicle is synthesized.

Furthermore, when the scanned vehicle appearance image data is still poor after the optimization process, the following processing is performed:

Based on an established model database containing a large number of complete vehicle appearances of different models, vehicle model data of the same model is queried from the model database according to the model of the current vehicle detected for replacement; or,

Check whether the current vehicle has historical scan data. If so, select the corresponding vehicle model data from the historical scan data for replacement.

Furthermore, the optimization process also includes: dynamically and in real time scanning the appearance image of the vehicle, acquiring a large amount of scanning data, establishing an artificial intelligence model, learning and understanding the defects of the imaging device during the imaging process and improving the working parameters of the imaging device.

Furthermore, establishing a unified three-dimensional coordinate system based on the working environment includes: determining a point within the structural layout as the coordinate origin based on the structural layout of the working environment where the vehicle is located, and determining the type and direction of the coordinate system; measuring and calibrating the positions of the structures within the environmental structure and the vehicle positions within the parking area, and determining their coordinate points relative to the coordinate origin.

Furthermore, the image data actually scanned and the position data measured and calibrated are corrected, and the correction includes:

The measured and calibrated position coordinate points of the structure and the vehicle relative to the origin are used as a reference, and compared with the coordinate points of the image data obtained by the actual scanning of the imaging equipment relative to the origin to calibrate the imaging accuracy error. If an error exists, the error range is preset in the imaging data and corrected.

Furthermore, the motion scanning includes mobile scanning, rotation scanning or a combination of the two, and the mobile scanning includes mobile scanning in the horizontal direction or the vertical direction.

Furthermore, when scanning different positions of the vehicle, the spatial position and posture of the imaging device are controlled and adjusted so that the incident source is close to the normal direction of the vehicle surface, wherein the scanning of the vehicle is based on the coordination of global scanning and local scanning.

Furthermore, the imaging device includes one or more of a laser radar, a visual sensor, an image sensor, a distance sensor, a millimeter wave sensor, an ultrasonic sensor, and a thermal imaging sensor, and performs fusion processing on image data collected by different types of imaging devices.

Another aspect of the present invention further provides a vehicle three-dimensional scanning imaging system, comprising:

The coordinate system establishment module builds a unified three-dimensional coordinate system based on the structural layout in the working environment;

An image acquisition module, used to obtain the appearance image data of the vehicle, wherein the appearance image data is obtained by scanning along the outer side of the vehicle parked in the parking area by using a plurality of imaging devices;

The image generation module is used to uniformly present the appearance image data of the vehicle acquired at different times in the three-dimensional coordinate system and perform data aggregation to form complete three-dimensional image data of the vehicle.

Furthermore, the system further comprises a data optimization module for judging the quality of the scanned shape image data and optimizing the data with quality defects, wherein the data optimization module comprises:

An image filling unit performs fuzzy filling processing on blank image areas appearing in the scanned vehicle appearance image data;

The mirror processing unit, based on the symmetry of the vehicle, when detecting that the image data obtained by scanning on one side is bad, performs mirror processing on the qualified image data obtained by scanning on the other side to replace it;

A vehicle model replacement unit, comprising a model database storing complete appearance features of different vehicle models and storing historical scan data of vehicles, and when the scanned imaging data of the current vehicle is still bad after being processed by the image filling unit or the mirror processing unit, the model data of the same model vehicle in the model database is replaced, or corresponding qualified model data is selected from the historical scan data of the vehicle for replacement;

The intelligent model building unit accumulates the appearance image data of the vehicle obtained through dynamic real-time scanning and builds an artificial intelligence model that includes the appearance features of the vehicle.

Compared with the prior art, the vehicle three-dimensional scanning imaging method and system provided by the present invention has the following technical effects:

1. The motion scanning of multiple types of sensors improves the resolution, three-dimensional measurement accuracy and imaging quality of imaging data.

2. By adjusting the emission angle of the sensor's signal source so that it can approach or coincide with the normal direction of the vehicle's shape, the collected image data is clearer and the problem of dark mirrors is weakened.

3. Low cost, reaching or exceeding the imaging accuracy and quality of similar high-beam LiDARs at a much lower cost.

4. Through the mutual coordination of different data optimization methods, different types of optimization are carried out according to different situations to form the best scanning imaging effect.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic flow chart of a vehicle three-dimensional scanning imaging method according to an embodiment of the present invention.

FIG. 2 is a schematic diagram of a vehicle three-dimensional scanning imaging system according to an embodiment of the present invention.

FIG. 3 is a schematic diagram showing the arrangement of an imaging device according to an embodiment of the present invention.

DETAILED DESCRIPTION

The following examples are only used to more clearly illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. For example, certain words are used in the specification and claims to refer to specific components. Those skilled in the art should understand that hardware or software manufacturers may use different nouns to refer to the same component. This specification and claims do not use the difference in name as a way to distinguish components, but use the difference in function of the components as the criterion for distinction. The subsequent description of the specification is a preferred embodiment of the present invention, but the description is for the purpose of illustrating the general principles of the present invention, and is not intended to limit the scope of the present invention. The scope of protection of the present invention shall be determined by the definition of the attached claims.

The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

An embodiment of the present invention provides a vehicle three-dimensional scanning imaging method, which is used for obtaining the vehicle's appearance features during self-service car washing, wherein the self-service car washing can be carried out in an indoor or outdoor car wash room, or in other specific environmental scenarios, a track is installed and fixed on the periphery of the parking area for vehicle washing, the track can be an elliptical track surrounding the vehicle, or one or more straight tracks arranged parallel or non-parallel along the length of the vehicle or in other directions, a motion terminal that can slide along the track is arranged on the track, the motion terminal can be a robot or a module that only has simple functions such as imaging, and the motion terminal has multiple and different types of sensors to perform all-round motion scanning of the vehicle's appearance features.

As shown in FIG1 , the imaging method comprises the following steps:

Step S11, establishing a unified three-dimensional coordinate system according to the current environment structure where the vehicle is located;

Among them, establishing a three-dimensional coordinate system includes:

According to the structural layout of the working environment where the vehicle is located, a certain point in the structural layout is determined as the coordinate origin, and the type and direction of the coordinate system are determined at the same time; taking the car wash room as an example, according to the size of the car wash room, the spatial structural layout, the parking area and other factors, a certain point in the space is selected as the coordinate origin, and then the direction of the coordinate axis and the coordinate type are determined to establish a three-dimensional coordinate system, such as a three-dimensional Cartesian coordinate system. After the unified coordinate system is established, the position data of all subsequent scanned image points will be calculated and judged in this coordinate system.

The structures in the environment structure and the vehicle positions in the parking area are measured and calibrated to determine their coordinate points relative to the coordinate origin. After the coordinate system is selected, all static structures in the current environment structure and subsequent vehicles entering the car wash room have corresponding position coordinate points. The selected structures may include tracks on the ground, lighting on the wall, wall lines, parking lines, etc.

After the structures and parking positions in the car wash are calibrated, the image point position data actually scanned by the imaging device will be obtained later. The coordinate points of the structures and vehicles relative to the origin measured and calibrated manually will be used as a reference and compared with the coordinate points of a certain image point position data obtained by the actual scanning of the imaging device relative to the origin. When there is an error between the two, the error range will be preset in the imaging data for correction. For example, the imaging coordinate point can be adjusted by adjusting the rotation axis of the imaging device. Such continuous corrections can keep the coordinate points of the two as consistent as possible, thereby improving the accuracy of image recognition.

Step S12: scanning the vehicles parked in the parking area at different times by using an imaging device to obtain the appearance image data of the vehicles and present them uniformly in the unified coordinate system;

Among them, motion scanning includes mobile scanning, rotation scanning or a combination of the two. Mobile scanning includes horizontal scanning, such as scanning when sliding linearly along the length direction of the vehicle or circular or arc scanning around the vehicle body, and vertical scanning, such as scanning when sliding linearly in the height direction of the vehicle, and can also be oblique scanning. The motion trajectory path of the imaging device is determined by the local computer or the background server after calculation, and its motion route is fixed and accurate. During the mobile scanning process, the imaging angle can also be rotated at the same time. The image data formed during the movement is essentially point cloud data. In this embodiment, the imaging device scans and images in real time during the movement process. On the one hand, the imaging device is moving, and its position will change in the coordinate system. The imaging data obtained is also determined based on the real-time position relationship of the imaging device. The relationship between the motion displacement of the imaging device and time, and the position relationship of the imaging device relative to the vehicle at a certain moment, are mutually corresponding. Since the displacement of the imaging device is relatively certain and clear, the vehicle imaging data obtained is also relatively clear. The imaging device can obtain point cloud data of different parts of the vehicle from different angles at any time, and accumulate and superimpose the point cloud data, which significantly improves the resolution of the imaging aggregation data and the imaging quality. The imaging resolution is reduced from about 10 cm of the original high-end laser radar to less than 1 cm, and the imaging accuracy is reduced from 2 cm of the original high-end laser radar to less than 1 cm.

The acquisition of vehicle appearance features is carried out through the cooperation of multiple different types of imaging devices. The imaging devices in this embodiment include radar (such as lidar, ultrasonic radar, millimeter wave radar, etc.), image sensors (such as high-definition cameras, binocular or multi-eye cameras, TOF cameras, thermal imaging cameras, etc.), and one or more distance sensors. It is preferred to use different types of imaging devices to collect data independently at the same time. Each sensor has its own unique advantages. After the image data collected by different types of sensors are fused and processed, data redundancy will be formed, and better image feature points can be selected from them, reducing the defects of single-type sensor imaging data that are prone to missing or poor imaging quality of a specific part of the vehicle, thereby more completely capturing the complete features of the vehicle and overcoming the problem of poor imaging quality on the dark mirror surface of the vehicle.

The machine can be one, two or more than two. There is a gap between the imaging sensor and the vehicle body. It is better to install the imaging sensor at a height higher than the highest point of the vehicle body to avoid missing the imaging data on the roof and to obtain a better imaging angle when imaging other vehicle positions. In the prior art, the incident source of the imaging sensor is vertically incident in the top and side view directions of the vehicle, and the imaging data cannot be captured in the vertical parts such as the front and rear of the vehicle. The present invention controls the sensor to adjust different spatial postures when scanning at different positions of the vehicle body. For example, the scanning following the shape of the car ensures that the incident source is as close to the normal of the car as possible. By controlling the running trajectory of the sensor, the scanning following the shape of the vehicle can be completed. Specifically, the imaging sensor is controlled to continuously move and/or rotate, and at least at a certain moment, the incident angle of the radiation signal generated by one or more imaging sensors coincides with or is close to the normal of the vehicle surface. Taking sensors that actively transmit signals, such as radar, distance sensors, and TOF cameras, as an example, the incident signal is incident from the normal direction, and its reflected signal returns along the normal line. Because it is nearly vertically incident on the surface of the object, the signal reflection performance is the strongest, and there is the greatest possibility of obtaining high-quality imaging data and ensuring high accuracy and high confidence of the data, which is specifically manifested in significantly improved available resolution, three-dimensional measurement accuracy, and significant improvement of dark mirror problems; it is convenient for the precise planning of subsequent cleaning trajectories and air-drying trajectories. In addition, the imaging sensor is in a rotating and moving form. Even if the signal source at a certain moment is not close to the normal line of the vehicle shape, it will be close to the normal line of the vehicle shape or coincide with the normal line of the vehicle shape at the next moment or other moments. The above method overcomes the core problem of the limited incident angle of the imaging sensor, and with the computer algorithm, it significantly weakens the dark mirror problem, significantly improves the imaging coverage and imaging quality of the vehicle body, and significantly improves the three-dimensional measurement accuracy.

In this embodiment, the image acquisition of the imaging device adopts a global scanning and local scanning method. Specifically, according to the division of labor of the nature of the work, it can be divided into a main sensor and a local sensor. The main sensor is responsible for the overall scanning of the vehicle data, and the local sensor is responsible for scanning and imaging the area where the main sensor cannot accurately obtain data. The main sensor and the local sensor scan and image at the same time. The overall data is based on the scanning data of the main sensor. The main sensor must ensure the largest possible scanning coverage area. The local sensor is responsible for scanning data at special locations such as rearview mirrors and bumpers. When arranging, the installation height of the main sensor on the machine is usually relatively high, such as being located on the top of a moving machine to obtain a relatively good scanning area. The arrangement position of the local sensor is mainly personalized according to the part to be scanned, such as being arranged at a position such as a rearview mirror. The local sensor can also perform motion control (including movement, rotation, etc.).

In addition, after obtaining the vehicle's appearance image data, a filtering algorithm (such as a Kalman filtering algorithm) is used to remove imaging data noise and retain high-quality data.

Step S13: Aggregate the appearance image data presented in the three-dimensional coordinate system to form complete three-dimensional image data of the vehicle.

Since the imaging device is in real time during the process of motion, it is necessary to aggregate the vehicle appearance image data obtained by the sensor from different angles at different times in a three-dimensional coordinate system, and finally form a complete vehicle three-dimensional image data, including the vehicle appearance and the position coordinates of each image point. Based on the formed vehicle three-dimensional image data, the local computer or backend server further calculates the cleaning path and air-drying path for the vehicle.

During image acquisition, the vehicle shape data obtained by the imaging device often has missing or other abnormal problems, resulting in poor imaging effects, especially in dark mirrors (especially black mirrors) and car glass. To solve this problem, the acquired data needs to be further optimized to achieve good imaging effects.

In this embodiment, the optimization processing of the acquired appearance image data includes the following methods:

Method 1: judge the continuity of the vehicle appearance image data and perform fuzzy filling processing on the blank image area.

The vehicle shape has obvious continuity characteristics. The continuity of the vehicle shape image data is used as the basis for judging whether the imaging data is complete. Among them, the continuity judgment of the vehicle can be judged by outlier detection algorithm, region growth or vehicle feature region prior knowledge algorithm. Outlier detection is the process of finding out its behavior is different from the expected object, including: outlier detection based on statistical methods, outlier detection based on distance, and outlier detection based on density. Taking the distance-based outlier detection method as an example, it considers the neighborhood of the object with a given radius. If there are not enough other points in its neighborhood, it is considered an outlier. The data that is obviously inconsistent with the law is removed through the outlier test. The region growing algorithm is to merge pixels with similar properties together, segment the connected areas that do not have the same characteristics, and provide good boundary information and segmentation results. Exemplarily, the scanned point cloud data connects adjacent points to form a grid, calculates the normal direction of the grid, and the point cloud grids with normals in the same direction are likely to belong to the same object. The prior knowledge of vehicle feature regions is to first separate the properties of different surface regions of the vehicle through feature models. For example, if one of the regions is the front engine cover, it is judged that there is no possibility of a large hole in this region and it needs to be repaired. When it is judged that there are some discontinuous image data of the vehicle, such as hollow or blank areas, fuzzy filling processing is performed. Fuzzy filling can automatically generate images for blank areas based on interpolation filling algorithms.

Method 2: Based on the symmetry of the vehicle, when it is detected that the image data obtained by scanning one side is bad, the qualified image data obtained by scanning the other side is mirrored, and finally the complete three-dimensional image data of the vehicle is synthesized.

Almost all vehicles on the market are symmetrically designed. For example, the exterior rearview mirror of a car is unlikely to exist in isolation. Based on the above characteristics, when the scan on one side is poor, the scan result on the other symmetrical side is used for mirror replacement. Specifically, the sensor motion scan is used to determine whether the data is missing. If a part of the data is seriously missing, the scan result on the other side corresponding to the missing part is found based on the center axis of the vehicle. In actual operation, the imaging of the wheel hub data is theoretically the best. In this way, it is very convenient to use the contour data and coordinates on both sides as a reference, establish a symmetry plane in the coordinate system, and perform mirror replacement of the left and right image data, and finally generate a complete three-dimensional image data.

Method 3: Establish a model database containing the complete appearance of vehicles of different models, construct a scene data model of structures and machinery and equipment in the current environment structure, locate the relative position of the vehicle in the current environment structure when scanning the vehicle, and when the vehicle scanning imaging is still poor after the optimization processing, replace it with the same type of vehicle model in the model database at the current position of the vehicle to be cleaned. In addition, the historical scanning data of the vehicle can also be stored. The stored historical scanning data is at least data that reflects the complete appearance characteristics of the vehicle, preferably data of the complete appearance characteristics formed from the most recent scan of the vehicle. When the vehicle has multiple scans, only one of the qualified scan data can be retained, and the best scan data can be used to replace the previous scan data.

Based on the actual vehicle appearance, when the imaging quality cannot be optimized or fails to meet the corresponding standards after optimization, the appearance data model of the same model of vehicle in the system database is directly replaced with the currently scanned data, or the corresponding data model is queried from the historical scan data of the vehicle that has been scanned before. In this way, the background server directly uses the model data of the same type of vehicle in the model database or the model data in the historical scan data as the basis, and plans the cleaning path and the air-drying path based on the queried vehicle model data.

The above three methods realize the optimization of vehicle scanning data from different directions, among which the above three methods can be carried out separately and independently, that is, different optimization methods can be directly selected according to different situations, or they can be coordinated with each other. For example, after completing the imaging scan, the optimization of method one is given priority, that is, the continuity of the vehicle appearance image features is judged. If there are some discontinuous parts, fuzzy filling processing is directly performed. After the processing of method one, most of the scanning results can be basically processed; however, when the discontinuous parts are large, or the scanning quality of some areas is too poor, there are many faults, and fuzzy filling cannot be performed, the symmetrical optimization method in method two can be used. Of course, method one can also be processed at the same time in method two, such as first fuzzy filling a certain part of the vehicle side with relatively good scanning results, and then the image data of the other side is obtained based on the mirror image of the side; if after the processing of methods one and two, it is still impossible to obtain good imaging data, then consider selecting the same type of vehicle model from the database or historical scanning data in method three for equivalent replacement.

As a preferred embodiment of the present invention, an artificial intelligence model is constructed through big data to continuously learn the advantages and disadvantages of different imaging devices or imaging data. The appearance image of the vehicle is dynamically scanned in real time to obtain a large amount of scanning data, and an artificial intelligence model is established to learn and understand the data defects in the machine imaging process and to improve and accurately predict them.

During the scanning and imaging process, some imaging effects may be poor to a greater or lesser extent. These conditions may be related to the environment, such as light, temperature and humidity, or to the scanning device itself, such as accuracy, angle, device parameters, or to the algorithm itself, such as filter parameters or filter models. By continuously scanning to obtain the vehicle's appearance image data, a large amount of scanning data can be obtained. Based on these scanning data and results, an artificial intelligence model is established. Through continuous learning and understanding of the model, the defects of the imaging device or imaging data can be discovered. According to the results, a set of parameter values with a more accurate recognition rate can be statistically calculated. Through the accumulation of big data, the parameters can be continuously adjusted and improved, and ultimately more accurate scanning results can be predicted.

A vehicle three-dimensional scanning and imaging method disclosed in an embodiment of the present invention significantly improves imaging accuracy and imaging quality through multi-sensor motion scanning within the same coordinate system, and at the same time combines multiple optimization processing methods to further improve imaging quality, especially in dark mirrors and vehicle glass parts, the imaging effect is better.

2 , corresponding to the above embodiment, another embodiment of the present invention provides a vehicle three-dimensional scanning imaging system, which includes: a coordinate system establishment module, an image acquisition module, and an image generation module.

The coordinate system establishment module builds a unified three-dimensional coordinate system based on the structural layout of the vehicle's operating environment;

The image acquisition module is used to obtain the appearance image data of the vehicle, wherein the image acquisition module obtains the image data by moving and scanning along the outer side of the vehicle parked in the parking area through a plurality of different types of imaging devices set up; FIG3 is a schematic diagram of the arrangement of the imaging device in an embodiment of the present invention, and exemplarily, the imaging device moves and scans in the horizontal direction along with the machines A and B on the slide rails on both sides.

The image generation module is used to uniformly present the appearance image data of the vehicle acquired at different times in the three-dimensional coordinate system and perform data aggregation to form complete three-dimensional image data of the vehicle.

In order to improve the imaging quality, the present invention is also provided with a data optimization module to determine the vehicle appearance image data with quality defects for further optimization, wherein the data optimization module includes: an image filling unit, a mirror processing unit and a vehicle model replacement unit.

The image filling unit performs fuzzy filling processing on the blank image area appearing in the scanned vehicle appearance image data.

The mirror processing unit, based on the symmetry characteristics of the vehicle, when detecting that the image data obtained by scanning on one side is bad, performs mirror processing on the qualified image data obtained by scanning on the other side to replace it; wherein, the mirror processing unit can be used simultaneously with the graphic filling unit.

The vehicle model replacement unit includes a model database storing complete appearance features of different vehicle models and storing historical scanning data of the vehicle. When the scanning imaging data of the current vehicle is still poor after being processed by the image filling unit or the mirror processing unit, the model data of the same model vehicle in the model database is replaced, or qualified model data is selected from the historical scanning data of the vehicle for replacement.

Of course, those skilled in the art should be able to understand that the various functional units in the above-mentioned data optimization module can be performed independently of each other or in cooperation with each other, but considering that even the same model may have minor modifications or changes, the image filling unit is used for processing first, followed by the mirror processing unit, and finally the vehicle model replacement unit.

Due to continuous scanning and imaging, big data accumulation is formed. The system of the present invention is also provided with a model building unit to accumulate the vehicle's appearance image data obtained by dynamic real-time scanning of the machine equipment, and establish an artificial intelligence model that includes the complete appearance features of the vehicle. By continuously learning the advantages and disadvantages of different imaging devices, the parameters of the imaging device are improved or the filter parameters are adjusted, thereby improving the accuracy and scanning efficiency of image scanning.

The vehicle three-dimensional scanning and imaging system in the embodiment of the present invention has the same technical effect as the vehicle three-dimensional scanning and imaging method described in the previous embodiment, so it will not be described in detail here.

It is worth noting that the above is only a preferred embodiment of the present invention, and does not limit the scope of patent protection of the present invention. The present invention can also improve the materials and structures of the above-mentioned various parts, or replace them with technical equivalents. Therefore, all equivalent structural changes made by using the description and illustrations of the present invention, or directly or indirectly applied to other related technical fields are also included in the scope of the present invention.

Claims (9)

1. A vehicle three-dimensional scanning imaging method, characterized in that: The method comprises: Establish a unified three-dimensional coordinate system based on the working environment of the vehicle; Performing motion scanning on vehicles parked in the parking area by using an imaging device to obtain image data of the vehicle's appearance at different times and presenting them uniformly in the unified three-dimensional coordinate system; Aggregating the appearance image data presented in the three-dimensional coordinate system to form complete three-dimensional image data of the vehicle; The method comprises: performing motion scanning on the vehicles parked in the parking area by an imaging device to obtain the appearance image data of the vehicles at different times and uniformly presenting them in the unified three-dimensional coordinate system, including: judging the point cloud quality of the scanned appearance image data and optimizing the point cloud with quality defects; The optimization process includes: judging the continuity of the vehicle shape image data, and performing fuzzy filling processing on the blank image area; when the discontinuous part is large, or the scanning quality of some areas is too poor, there are many faults, and fuzzy filling cannot be performed, according to the symmetry characteristics of the vehicle, when it is detected that the image data obtained by scanning one side of the vehicle is poor, the qualified image data obtained by scanning the other side is mirrored, and finally the complete three-dimensional image data of the vehicle is synthesized; When the scanned vehicle's appearance image data is still poor after the optimization processing, the following processing is performed: based on the established model database containing a large number of complete vehicle appearances of different models, the vehicle model data of the same model is queried from the model database according to the model of the current vehicle detected for replacement; or, whether the current vehicle has historical scanning data, if so, the corresponding vehicle model data is selected from the historical scanning data for replacement. 2. The imaging method according to claim 1, characterized in that The optimization process also includes: dynamically and in real time scanning the appearance image of the vehicle, acquiring a large amount of scanning data, establishing an artificial intelligence model, learning and understanding the defects of the imaging device during the imaging process, and improving the working parameters of the imaging device. 3. The imaging method according to claim 1, characterized in that: Establishing a unified three-dimensional coordinate system according to the working environment includes: determining a point in the structural layout as the coordinate origin based on the structural layout of the working environment where the vehicle is located, and determining the type and direction of the coordinate system; measuring and calibrating the positions of the structures in the environmental structure and the vehicle positions in the parking area to determine their coordinate points relative to the coordinate origin. 4. The imaging method according to claim 3, characterized in that: During the scanning process, the actual scanned image data and the measured and calibrated position data are corrected, and the correction includes: The measured and calibrated position coordinate points of the structure and the vehicle relative to the origin are used as a reference, and compared with the coordinate points of the image data obtained by the actual scanning of the imaging equipment relative to the origin to calibrate the imaging accuracy error. If an error exists, the error range is preset in the imaging data and corrected. 5. The imaging method according to claim 1, characterized in that: The motion scanning includes mobile scanning, rotation scanning or a combination of the two, and the mobile scanning includes mobile scanning in the horizontal direction, vertical direction or oblique direction. 6. The imaging method according to claim 5, characterized in that: When scanning different positions of the vehicle, the position and posture of the imaging device are controlled and adjusted so that the incident source is close to the normal direction of the vehicle surface. The scanning of the vehicle is based on the coordination of global scanning and local scanning. 7. The imaging method according to claim 1 or 6, characterized in that: The imaging device includes one or more of a laser radar, a visual sensor, an image sensor, a distance sensor, a millimeter wave sensor, an ultrasonic sensor, and a thermal imaging sensor, and performs fusion processing on image data collected by different types of imaging devices. 8. A vehicle three-dimensional scanning imaging system, characterized in that the system comprises: A coordinate system building module is used to build a unified three-dimensional coordinate system according to the structural layout in the working environment; An image acquisition module, used to obtain the appearance image data of the vehicle, wherein the appearance image data is obtained by scanning along the outer side of the vehicle parked in the parking area by using a plurality of imaging devices; An image generation module, used to uniformly present the appearance image data of the vehicle acquired at different times in the three-dimensional coordinate system and perform data aggregation to form complete three-dimensional image data of the vehicle; The system further includes a data optimization module for judging the quality of the scanned shape image data and optimizing the data with quality defects, wherein the data optimization module includes: An image filling unit, used for performing fuzzy filling processing on blank image areas appearing in the scanned vehicle appearance image data; The mirror processing unit is used to replace the qualified image data obtained by scanning on the other side by mirror processing according to the symmetry characteristics of the vehicle when the image data obtained by scanning on one side is detected to be bad when the discontinuous parts are large, or the scanning quality of some areas is too poor, there are many faults, and fuzzy filling cannot be performed; The vehicle model replacement unit includes a model database storing complete appearance features of different vehicle models and storing historical scanning data of vehicles. When the scanning imaging data of the current vehicle is still poor after being processed by the image filling unit and the mirror processing unit, the model data of the same model vehicle in the model database is replaced, or the corresponding qualified model data is selected from the historical scanning data of the vehicle for replacement. 9. The system as described in claim 8 is characterized in that the data optimization module also includes an intelligent model building unit, which is used to accumulate the vehicle's appearance image data obtained by dynamic real-time scanning and build an artificial intelligence model containing vehicle appearance features.