CN117429413A - Empty space detection method and device, intelligent automobile and readable storage medium - Google Patents

Abstract

The invention relates to the field of automatic driving, and discloses a method and a device for detecting an empty space, an intelligent automobile and a readable storage medium. The method comprises the following steps: acquiring original distance data returned by the detection devices in all directions on the vehicle to obtain dynamic coordinates of all detected objects; determining a real-time boundary distance between the vehicle and each detected object according to the dynamic coordinates; determining a distance mutation value according to the difference value between the boundary distance at the current moment and the boundary distance at the last moment, and determining whether to generate a detection contour according to the distance mutation value; and determining the contour width and the contour depth according to the generated detection contour, and determining that the empty vehicle position is detected if the contour width and the contour depth reach preset values. The vehicle can determine the parking space through the outline, and the vehicle is more accurate.

Description

Empty parking space detection method, device, smart car and readable storage medium

Technical field

The invention relates to the field of automatic driving, and in particular to an empty parking space detection method, device, smart car and readable storage medium.

Background technique

The APA parking system is an automatic parking assistance system that uses the vehicle's surround-view camera, ultrasonic and other sensors to detect the environment around the vehicle, automatically find a suitable parking space, and automatically complete the parking operation.

In APA parking, ultrasonic sensors are usually used to assist visual recognition of empty parking spaces. When the car drives through the empty parking space at low speed, the ultrasonic sensor installed on the side of the car will detect the distance. When there are obstacles or vehicles on both sides of the empty parking space, the ultrasonic sensor detects The distance will first become smaller, then larger, and then smaller again. The system detects this process and obtains empty parking space information based on vehicle positioning information, and hands the obtained empty parking space to the subsequent planning module for automatic parking. control.

At present, there is a certain error in the distance information returned by the ultrasonic sensor, and it is greatly affected by the surrounding environment: the ultrasonic sensor has a beam angle, and the echo jump is unstable, resulting in early echo or lagging echo, thereby causing obstacles. Positioning error means that the detected parking space is inaccurate, or misjudgment or omission occurs.

Contents of the invention

In the first aspect, this application provides an empty parking space detection method, including:

Obtain the original distance data returned by the detection device in each direction on the vehicle and obtain the dynamic coordinates of each detection object;

Determine the real-time boundary distance between the vehicle and each detected object according to the dynamic coordinates;

Determine the distance mutation value based on the difference between the boundary distance at the current moment and the boundary distance at the previous moment, and determine whether to generate a detection contour based on the distance mutation value;

According to the generated detection contour, the contour width and contour depth are determined. If both the contour width and the contour depth reach the preset value, it is determined that an empty parking space has been detected.

Further, determining whether to generate a detection contour according to the distance mutation value includes:

Determine the change interval of the distance mutation value, and each change interval has a set weight value;

If the mutation direction of the distance mutation value is the same as the mutation direction in the last detection, then the weight values corresponding to the change intervals are accumulated according to the change interval where the distance mutation value is located. If the accumulated weight value reaches If the threshold is preset, the detection contour will be generated.

Further, the method also includes:

If the mutation direction of the distance mutation value is different from the mutation direction in the last detection, then the weight value is cleared to zero;

Return to the step of determining the real-time boundary distance between the vehicle and each detected object.

Further, the contour width and contour depth are determined based on the currently generated detection contour. If both the contour width and the contour depth reach a preset value, it is determined that an empty parking space is detected, including:

Calculate the distance between adjacent contours to obtain the contour distance;

If the contour distance is greater than the first preset distance, then the boundary distance of the area between the adjacent contours is used as the contour depth;

If the contour depth is greater than the second preset distance, it is determined that an empty parking space is detected.

Further, determining whether to generate a detection contour according to the distance mutation value includes:

Determine the mutation direction of the current distance mutation value. If the mutation direction is different from the mutation direction in the last detection, it is determined to generate a new contour.

Further, obtaining the original distance data returned by the detection device in each direction of the vehicle includes:

Obtain the object coordinates detected by the detection devices in the four directions of the vehicle's left front, right front, left rear, and right rear.

Further, the method also includes:

The contours generated by different detection devices in the same area are fused. If any detection device identifies that there is an empty parking space in the current area, it is determined that there is an empty parking space in the current area.

In a second aspect, this application also provides an empty parking space detection device, including:

A detection module, used to obtain the original distance data returned by the detection device in each direction on the vehicle and obtain the dynamic coordinates of each detection object;

A ranging module, used to determine the real-time boundary distance between the vehicle and each detected object according to the dynamic coordinates;

A contour generation module, configured to determine a distance mutation value based on the difference between the boundary distance at the current moment and the boundary distance at the previous moment, and determine whether to generate a detection contour based on the distance mutation value;

The judgment module is used to determine the contour width and contour depth according to the generated detection contour. If the contour width and the contour depth both reach a preset value, it is determined that an empty parking space has been detected.

In a third aspect, this application also provides a smart car, including a processor and a memory. The memory stores a computer program, and the computer program executes the empty parking space detection method when running on the processor.

In a fourth aspect, the present application also provides a readable storage medium that stores a computer program that executes the empty parking space detection method when running on a processor.

The invention discloses an empty parking space detection method, device, smart car and readable storage medium. The method includes: obtaining the original distance data returned by the detection device in each direction on the vehicle and obtaining the dynamic coordinates of each detection object; determining the boundary distance between the vehicle and each detection object according to the dynamic coordinates; The difference between the boundary distance at the current moment and the boundary distance at the previous moment determines the distance mutation value. According to the distance mutation value, it is determined whether to generate a detection contour; according to the currently generated contour, the contour width and contour depth are determined. If the contour When both the width and the contour depth reach the preset value, it is determined that an empty parking space has been detected. This allows vehicles to determine parking spaces through contours more accurately.

Description of the drawings

In order to explain the technical solutions of the present invention more clearly, the drawings required to be used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and therefore should not be interpreted as It is regarded as limiting the scope of protection of the present invention. In the various drawings, similar components are numbered similarly.

Figure 1 shows a schematic flow chart of an empty parking space detection method according to an embodiment of the present application;

Figure 2 shows a schematic diagram of a parking scene according to an embodiment of the present application;

Figure 3 shows a schematic diagram of the world coordinate system when a vehicle is parking according to an embodiment of the present application;

Figure 4 shows a schematic diagram of a detection profile according to an embodiment of the present application;

Figure 5 shows a schematic structural diagram of an empty parking space detection device according to an embodiment of the present application.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all the embodiments.

The components of the embodiments of the invention generally described and illustrated in the figures herein may be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of the invention provided in the appended drawings is not intended to limit the scope of the claimed invention, but rather to represent selected embodiments of the invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without any creative work fall within the scope of protection of the present invention.

Hereinafter, the terms "comprising", "having" and their cognates, which may be used in various embodiments of the present invention, are only intended to represent specific features, numbers, steps, operations, elements, components or combinations of the foregoing. and should not be understood as first excluding the presence of one or more other features, numbers, steps, operations, elements, components or combinations of the foregoing or adding one or more features, numbers, steps, operations, elements, components or the possibility of a combination of the foregoing.

In addition, the terms "first", "second", "third", etc. are only used to distinguish descriptions and shall not be understood as indicating or implying relative importance.

Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which various embodiments of this invention belong. Said terms (such as terms defined in commonly used dictionaries) will be interpreted to have the same meaning as the contextual meaning in the relevant technical field and will not be interpreted as having an idealized meaning or an overly formal meaning, Unless expressly limited in various embodiments of the invention.

This application is applied to the application scenario of obtaining the position of an empty parking space through automatic detection by a detection device when parking a vehicle. The dynamic coordinates of each detection object are obtained mainly by obtaining the original distance data returned by the detection device in each direction of the vehicle; According to the dynamic coordinates, the boundary distance between the vehicle and each detected object is determined; according to the difference between the boundary distance at the current moment and the boundary distance at the previous moment, a distance mutation value is determined, and according to the distance mutation value, it is determined whether Generate a detection outline; after generating the detection outline, determine the outline width and outline depth. If both the outline width and the outline depth reach a preset value, it is determined that an empty parking space has been detected. In this way, the location of the empty parking space is determined through the outline and detection is achieved.

Next, the technical solution of the present application will be described with specific embodiments.

Example 1

As shown in Figure 1, the empty parking space detection method in this embodiment includes:

Step S100, obtain the original distance data returned by the detection device in each direction on the vehicle and obtain the dynamic coordinates of each detection object;

The empty parking space detection method of this embodiment is obviously used in the parking process, in which multiple detection devices are installed on the vehicle. This detection device is mainly used to detect the distance between the obstacles around the vehicle and the vehicle. These detection devices can be Ultrasonic radar can also be a device such as a visual sensor.

These devices are arranged around the vehicle according to a certain distribution, as shown in Figure 2. In this parking scene, the vehicle 100 drives from right to left, and the four positions of the vehicle 100 include detectors at the left front, left rear, right front, and right rear. Device 200, the layout and number of detection devices 200 in Figure 2 are a feasible way. In addition to preventing detection at the four corners of the vehicle, corresponding detection devices can also be provided on the four sides of the vehicle.

At the same time, there are parking spaces and vehicles parked in the parking spaces on the right side of the vehicle 100. Through the layout of the detection device as shown in the figure, the surroundings of the vehicle can be obtained.

It can be seen that the vehicle 100 can detect the coordinate positions of surrounding obstacles through the above four detection devices. Among them, the vehicle is in motion, so the coordinates of surrounding obstacles will be dynamically updated due to the movement of the vehicle, so the obtained coordinates are dynamic coordinates. And it can be understood that the number of objects and coordinates seen is not just one, but multiple. By connecting these coordinates, an outline can be obtained.

Step S200: Determine the boundary distance between the vehicle and each detection object according to the dynamic coordinates.

In this embodiment, the midpoint of the rear wheel axle of the vehicle is used as the coordinates of the vehicle, and the boundary distance between each dynamic coordinate and the vehicle is calculated based on this.

Among them, when the vehicle is parking, in order to better determine the dynamic coordinates and boundary distance, the starting point of the vehicle will be the far point, and the center of the rear axle of the vehicle will be the vehicle coordinates to establish a world coordinate system. This coordinate system is shown in Figure 3 .

Among them, point P is the center of the rear axle of the vehicle, the position coordinate is (xoc, yoc), the heading angle of the vehicle is θ, FL, RL, FR, and RR are the ultrasonic radars of the left front, left rear, right front, and right rear respectively. The left front (right front) ), the lateral distances of the left rear (right rear) from the vehicle rear axle center point P are SNS1 and SNS2 respectively, and the longitudinal distances of the left front, right front, left rear and right rear from the vehicle rear axle center point P are SNS_Y.

Through the trigonometric formula, the coordinates of the four ultrasonic radars in the global coordinate system can be obtained:

FLx＝xoc+SNS2·cosθ-SNS_Y·sinθ

FLy＝yoc+SNS2·sinθ+SNS_Y·cosθ

RLx＝xoc-SNS1·cosθ-SNS_Y·sinθ

RLy＝yoc-SNS1·sinθ+SNS_Y·cosθ

FRx＝xoc+SNS2·cosθ+SNS_Y·sinθ

FRy＝yoc+SNS2·sinθ-SNS_Y·cosθ

RRx＝xoc-SNS1·cosθ+SNS_Y·sinθ

RRy＝yoc-SNS1·sinθ-SNS_Y·cosθ

Assume that in the current posture, the distances detected by the ultrasonic radar of the left front (FL), left rear (RL), right front (FR), and right rear (RR) are d0, d1, d2, and d3 respectively, then the currently detected obstacle coordinates They are:

(FLx-d0·sinθ, FLy+d0·cosθ),

(RLx-d1·sinθ, RLy+d1·cosθ),

(FRx+d2·sinθ, FRy-d2·cosθ),

(RRx+d3·sinθ, RRy-d3·cosθ).

Step S300: Determine a distance mutation value based on the difference between the boundary distance at the current moment and the boundary distance at the previous moment, and determine whether to generate a detection contour based on the distance mutation value.

The mutation value refers to the difference between the boundary distance at the previous moment and the boundary distance at this moment.

The boundary distance will change with the movement of the vehicle. There are generally 6 states. For example, the first preset distance, the second preset distance and the third preset distance are set from small to large. When the return distance is stable, the detected boundary distance If it is greater than the third preset distance and the change range of the mutation value is less than the first preset distance, it is a vacancy state.

If the returned boundary distance is stable and less than the third preset distance, it is an obstacle state.

If the boundary distance becomes significantly larger and the mutation value is within the first interval, it is a small upward trend state. If the mutation value is within the second interval, it is a large upward trend state.

If the boundary distance becomes significantly smaller and the mutation value is within the first interval, it is a small downward trend state. If the mutation value is within the second interval, it is a large downward trend state.

The above-mentioned first preset distance may be 20 to 40 cm, the second preset distance may be 80 to 100 cm, and the third preset distance may be 2000 to 3500 cm, then the first interval is determined by the first preset distance. value and the second preset value. The second interval can be composed of the second preset value and the third preset value. The interval changes with the change of the three preset values, as long as the range of the first interval is smaller than the second preset value. The range of the interval is enough.

Among them, the above four trend states are unstable states.

When it is in the vacancy state and obstacle state, it is a stable state. If a contour has been generated in this state, the contour can continue to be extended without generating a detection contour.

For the subsequent four upward trends and downward trends, it means that the distance between the surrounding objects and the vehicle has changed suddenly, and it may be necessary to generate detection contours to judge the surrounding environment.

To this end, a first weight is set when the mutation value is in the first interval, and a second weight is set when the mutation value is in the second interval. After each update of the boundary distance, the mutation value is calculated, and the mutation value is calculated according to the position of the mutation value. interval, and the weights are accumulated until the weight reaches a preset value, it is considered that a detection contour needs to be generated.

In addition, if the mutation direction of the mutation value is different from the last time, the weight value is cleared to zero.

It can be understood that this situation is generally caused by a complex surrounding obstacle environment. There are obstacles in a certain area and there are no obstacles in a certain area, so that the detection device will detect mutation values with different trends. In this case, it is impossible to form an effective detection method. Outline.

For example, the first weight is 1, the second weight is 5, and the preset weight value is 20. That is to say, if the detection mutation values for 4 consecutive times are all within the second interval, it is considered that a detection contour needs to be generated. And when the weight value reaches the preset value, the current state will also be judged as a stable state. In steady state, no new contours are generated.

One of the contours will contain the starting point coordinates and the end point coordinates of the contour. When the detection contour is generated, the coordinates of the obstacle detected at this time are the starting point coordinates of the detection contour.

Before generating the detection contour, determine whether the old contour width is smaller than the contour width threshold (such as less than 5cm). If it is smaller, it will be judged as interference data and the old contour will be deleted;

Before generating the detection contour, it is judged whether the current contour has the same characteristics as the previous contour (the same distance from the car body). If so, it means that interference data may be received in the middle, and the two contours are merged.

Step S400: After the detection contour is generated, the contour width and contour depth are determined. If both the contour width and the contour depth reach a preset value, it is determined that an empty parking space has been detected.

When generating a detection profile, the profile width and profile depth are also determined.

As shown in Figure 4, it is a contour image generated by a detection device, which corresponds to the parking lot scene in Figure 2.

The outline consists of line segments 1, 2, 3, 4, and 5. Line segments 1 and 2 represent the outline formed by the vehicle parked in the parking space in front of the vehicle in Figure 2. Line segments 4 and 5 represent the outline formed by the vehicle parked behind the vehicle. For vehicles in the parking space, line segment 3 represents the parking space stopper and other restraints that exist inside the parking space. It can be understood that according to the aforementioned contour generation rules, there are actually three groups of contours in Figure 4, that is, line segments 1 and 2 are a group, called the first contour, line segment 3 is the second contour, and line segments 4 and 5 are the third contour. .

Obviously, the first and third contours indicate that there are some physical obstacles in this part of the space, and vehicles cannot pass through them. The second contour indicates that no physical obstacles are detected here, and vehicles can pass through.

Therefore, the contour width in this step refers to the distance between the first contour and the third contour, which is determined to be an area with obstacles, and the area where the second contour is located is determined to be a vacancy state.

The contour width is the length represented by the line segment D marked in Figure 4, and the contour depth refers to the boundary distance detected in the space where the contour width is located, that is, the boundary distance detected by the detection device 200 in the area where these contours are located. .

It can be understood that generally speaking, the depth of the contour of an empty parking space will be greater than the width or length of the vehicle, and therefore will also be greater than the distance between the contours on both sides and the vehicle. In other words, the space of the empty parking space will comply with the above-mentioned vacancy status.

Among them, the contour width must be at least larger than the width of the vehicle itself to be enough for parking, and the contour depth must be at least larger than the vehicle width. Because parking spaces are generally longitudinal parking spaces or transverse parking spaces. If it is a longitudinal parking space, the outline width may be smaller than the vehicle length but must be greater than the vehicle width. At the same time, the outline depth must be greater than the vehicle length. Similarly, if it is a transverse parking space, the outline width must be greater than the vehicle length. , and the profile depth also needs to be greater than the vehicle width. That is to say, only the contour width and depth that meet the above conditions can be judged as detecting an empty parking space.

In addition, combined with the layout and driving direction of the detection device 200 in Figure 2, it can be seen that for the same area, there are often multiple detection devices 200 scanning, and the data of each detection device can form an outline according to the above steps. Therefore, for For these contours, this embodiment will also perform contour fusion operations.

Specifically, according to the maximizing outline scheme, the outlines formed by the front and rear of the same side of the vehicle are merged, which can further eliminate the interference of the surrounding environment and obtain a more accurate parking space outline. That is, as long as the outline of one detection device in a certain area is determined to be a stable vacancy state, no matter what state the other detection device determines, this area will be set to the vacancy state; at the same time, it will be recorded and so on, traversing all dynamic coordinate points , which can fuse the contours of the same side of the vehicle.

Specifically, traversing the area where the vehicle has traveled, there are only three situations in which the contours detected by the front detection device A and the contours obtained by the rear detection device B are merged, as follows:

There are no contours of A or B in this area: this area is determined to be in a transitional state and is not processed as an unknown contour.

There is only one outline of A and B in this area: adding this outline feature after merging makes the outline clearer.

There are outlines for both A and B in this area: after merging, the distance between the intersection area outline and the central axis takes the value that is farther from the car body.

For example, in the area where the X coordinate is 100cm to 150cm, A and B respectively obtain the following contour information:

A profile: starting point coordinates (100, 130), end point coordinates (150, 150) distance from the vehicle center axis 130cm ~ 150cm.

B contour: starting point coordinates (100, 150), end point coordinates (150, 130) distance from the vehicle center axis 150cm ~ 130cm.

The merged outline is the starting point coordinate (100, 150) and the end point coordinate (150, 150) being 150cm away from the vehicle center axis.

At the same time, each detection device often has a maximum range. For example, when the parking space boundary is too far from the vehicle, the actual effective detection range will be too narrow, and the current contour cannot be used for parking space judgment. First check the difference between the contour depth and the boundary to determine whether the depth is less than the vehicle width threshold. If it is less than the threshold, continue to wait for subsequent detection.

For example, the maximum range of a detection device is 4m. If the parking space boundary distance of a certain contour is greater than 3m, the effective detection depth of the maximum identified vacant space will be less than 1m, and this contour will not be output as a valid parking space.

After it is determined that there is a parking space, the parking space coordinates in the global coordinate system will be converted into the body coordinate system with the center of the rear wheel of the current vehicle as the origin to facilitate the parking operation of the vehicle.

If the global coordinate system parking space is (x0, y0), (x1, y1), (x2, y2), (x3, y3).

Then calculate the parking space coordinates in the body coordinate system:

((x0’, y0’), (x1’, y1’), (x2’, y2’), (x3’, y3’)) The method is as follows

x0’=(x0-xoc)·cos(θ)+(y0-yoc)·sin(θ)

y0’=(y0-yoc)·cos(θ)-(x0-xoc)·sin(θ)

x1’=(x1-xoc)·cos(θ)+(y1-yoc)·sin(θ)

y1’=(y1-yoc)·cos(θ)-(x1-xoc)·sin(θ)

x2’=(x2-xoc)·cos(θ)+(y2-yoc)·sin(θ)

y2’=(y2-yoc)·cos(θ)-(x2-xoc)·sin(θ)

x3’=(x3-xoc)·cos(θ)+(y3-yoc)·sin(θ)

y3’=(y3-yoc)·cos(θ)-(x3-xoc)·sin(θ)

Among them, θ in the above formula represents the heading angle of the vehicle when traveling, and yoc and xoc are the coordinates of the vehicle position in the global coordinate system.

This embodiment collects the detected obstacle distance and determines whether to generate a detection contour through the change of the boundary distance mutation value, generates the corresponding contour, and calculates the contour distance and contour depth, thereby helping the vehicle determine whether there is an idle parking space nearby. The method of this embodiment makes the formation of contours more stable, and takes into account the influence of mutation values, so that the generated contours are basically at the same distance from the vehicle. The generated contour results have more reference significance, and through merging, it can be further Remove interfering signals.

Example 2

As shown in Figure 5, this application also provides an empty parking space detection device, including:

The detection module 10 is used to obtain the original distance data returned by the detection device in each direction on the vehicle and obtain the dynamic coordinates of each detection object;

A ranging module 20 is used to determine the boundary distance between the vehicle and each detected object according to the dynamic coordinates;

The contour generation module 30 is configured to determine a distance mutation value based on the difference between the boundary distance at the current moment and the boundary distance at the previous moment, and determine whether to generate a detection contour based on the distance mutation value;

The judgment module 40 is used to determine the width and depth of the contour after generating the detection contour. If both the width of the contour and the depth of the contour reach a preset value, it is determined that an empty parking space has been detected.

This application also provides a smart car including a processor and a memory, the memory stores a computer program, and the computer program executes the empty parking space detection method when running on the processor.

This application also provides a readable storage medium, which stores a computer program, and the computer program executes the empty parking space detection method when running on a processor.

In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are only illustrative. For example, the flow charts and structural diagrams in the accompanying drawings show the possible implementation architecture and functions of the devices, methods and computer program products according to multiple embodiments of the present invention. and operations. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code that contains one or more components for implementing the specified logical function(s). Executable instructions. It should also be noted that, in alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two consecutive blocks may actually execute substantially in parallel, or they may sometimes execute in the reverse order, depending on the functionality involved. It will also be noted that each block in the structure diagrams and/or flowchart illustrations, and combinations of blocks in the structure diagrams and/or flowchart illustrations, can be configured with specialized hardware-based systems that perform the specified functions or actions. to be implemented, or may be implemented using a combination of dedicated hardware and computer instructions.

In addition, each functional module or unit in various embodiments of the present invention can be integrated together to form an independent part, each module can exist alone, or two or more modules can be integrated to form an independent part.

If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which can be a smart phone, a personal computer, a server, a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. .

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention. should be covered by the protection scope of the present invention.

Claims (10)

1. A method for detecting empty parking spaces, which is characterized by including: Obtain the original distance data returned by the detection device in each direction on the vehicle and obtain the dynamic coordinates of each detection object; Determine the real-time boundary distance between the vehicle and each detected object according to the dynamic coordinates; Determine the distance mutation value based on the difference between the boundary distance at the current moment and the boundary distance at the previous moment, and determine whether to generate a detection contour based on the distance mutation value; According to the generated detection contour, the contour width and contour depth are determined. If both the contour width and the contour depth reach the preset value, it is determined that an empty parking space has been detected. 2. The empty parking space detection method according to claim 1, characterized in that, determining whether to generate a detection contour according to the distance mutation value includes: Determine the change interval of the distance mutation value, and each change interval has a set weight value; If the mutation direction of the distance mutation value is the same as the mutation direction in the last detection, then the weight values corresponding to the change intervals are accumulated according to the change interval where the distance mutation value is located. If the accumulated weight value reaches If the threshold is preset, the detection contour will be generated. 3. The method for detecting empty parking spaces according to claim 2, further comprising: If the mutation direction of the distance mutation value is different from the mutation direction in the last detection, then the weight value is cleared to zero; Return to the step of determining the real-time boundary distance between the vehicle and each detected object. 4. The method for detecting empty parking spaces according to claim 1, wherein the contour width and contour depth are determined according to the generated detection contour. If both the contour width and the contour depth reach a preset value, then Confirm that empty parking spaces are detected, including: Calculate the distance between adjacent contours to obtain the contour distance; If the contour distance is greater than the first preset distance, then the boundary distance of the area between the adjacent contours is used as the contour depth; If the contour depth is greater than the second preset distance, it is determined that an empty parking space is detected. 5. The method for detecting empty parking spaces according to claim 1, wherein determining whether to generate a detection contour according to the distance mutation value includes: Determine the mutation direction of the current distance mutation value. If the mutation direction is different from the mutation direction in the last detection, it is determined to generate a new contour. 6. The empty parking space detection method according to claim 1, characterized in that said obtaining the original distance data returned by the detection device in each direction of the vehicle includes: Obtain the object coordinates detected by the detection devices in the four directions of the vehicle's left front, right front, left rear, and right rear. 7. The method for detecting empty parking spaces according to claim 1, further comprising: The contours generated by different detection devices in the same area are fused. If any detection device identifies that there is an empty parking space in the current area, it is determined that there is an empty parking space in the current area. 8. An empty parking space detection device, characterized in that it includes: The detection module is used to obtain the original distance data returned by the detection device in all directions on the vehicle and obtain the dynamic coordinates of each detection object; A ranging module, used to determine the real-time boundary distance between the vehicle and each detected object according to the dynamic coordinates; A contour generation module, configured to determine a distance mutation value based on the difference between the boundary distance at the current moment and the boundary distance at the previous moment, and determine whether to generate a detection contour based on the distance mutation value; The judgment module is used to determine the contour width and contour depth according to the generated detection contour. If the contour width and the contour depth both reach a preset value, it is determined that an empty parking space has been detected. 9. A smart car, characterized in that it includes a processor and a memory, the memory stores a computer program, and the computer program executes the method described in any one of claims 1 to 7 when running on the processor. Empty parking space detection method. 10. A computer-readable storage medium, characterized in that it stores a computer program that executes the empty parking space detection method according to any one of claims 1 to 7 when running on a processor.