CN112784705A - Vehicle side edge determining method and device - Google Patents
Vehicle side edge determining method and device Download PDFInfo
- Publication number
- CN112784705A CN112784705A CN202110004834.XA CN202110004834A CN112784705A CN 112784705 A CN112784705 A CN 112784705A CN 202110004834 A CN202110004834 A CN 202110004834A CN 112784705 A CN112784705 A CN 112784705A
- Authority
- CN
- China
- Prior art keywords
- output
- side edge
- vehicle
- vehicle image
- position information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the invention provides a method and a device for determining a vehicle side edge, wherein the method comprises the following steps: acquiring a vehicle image acquired by a camera and a pre-trained prediction model; the predictive model includes a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image; acquiring output values of a plurality of output items output by the prediction model to the vehicle image; determining whether a side edge exists according to the output value of the output item indicating whether the side edge of the vehicle exists; and if the existence of the side edge is determined, determining the position information of the side edge in the vehicle image according to the output value of the output item corresponding to the side edge determined to exist. Compared with the scheme of predicting the vehicle side edge through the laser radar, the embodiment of the invention can reduce the cost and improve the prediction accuracy.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a vehicle side edge determining method and a vehicle side edge determining apparatus.
Background
The automatic driving perception system of the vehicle can control automatic driving according to the local yaw angle of the vehicle in the driving process, and the existing automatic driving perception system generally uses sensors with depth information, such as laser radars and the like, to predict the local yaw angle of the vehicle. However, the cost of the lidar is high, and the target detection difficulty is high on the point cloud generated by the lidar, so that the prediction method is not commercially universal.
Disclosure of Invention
In view of the above, embodiments of the present invention have been made to provide a vehicle side edge determining method and a corresponding vehicle side edge determining apparatus that overcome or at least partially solve the above-described problems.
In order to solve the above problem, an embodiment of the present invention discloses a vehicle side edge determining method, including:
acquiring a vehicle image acquired by a camera and a pre-trained prediction model; the predictive model includes a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image;
acquiring output values of a plurality of output items output by the prediction model to the vehicle image;
determining whether a side edge exists according to the output value of the output item indicating whether the side edge of the vehicle exists;
and if the existence of the side edge is determined, determining the position information of the side edge in the vehicle image according to the output value of the output item corresponding to the side edge determined to exist.
Optionally, the determining whether the side edge exists according to an output value of an output item corresponding to whether the side edge of the vehicle exists includes:
normalizing the output values of the output items respectively representing whether the side edges of the vehicle exist or not;
determining whether a normalized output value greater than a preset threshold value exists;
and if the normalized output value greater than the preset threshold value exists, determining that the corresponding side edge exists.
Optionally, the method further comprises:
training the predictive model by:
obtaining a vehicle image sample and a predictive model, the predictive model comprising a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image;
obtaining output values of a plurality of output items output by the prediction model to the vehicle image sample;
calculating a first loss value for the output items respectively representing whether the side edges of the vehicle exist or not;
calculating a second loss value for the output items respectively representing the position information of the side edges in the vehicle image;
and training the prediction model according to the first loss value and the second loss value.
Optionally, the calculating a first loss value for the output items respectively indicating whether the side edges of the vehicle are present includes:
for the output items respectively representing whether the side edges of the vehicle exist or not, a first loss value is calculated by using a preset first loss function.
Optionally, the vehicle image sample has annotation information indicating whether a side edge exists; the calculating, for the output items respectively representing the position information of the side edges in the vehicle image, a second loss value including:
if the side edge is marked as existing, calculating a second loss value according to a corresponding output value for an output item corresponding to the side edge and representing the position information of the side edge in the vehicle image;
if a side edge is marked as not present, a second loss value of 0 is determined for the output item corresponding to the side edge and representing the position information of the side edge in the vehicle image.
Optionally, if a side edge is marked as present, calculating a second loss value according to a corresponding output value for an output item corresponding to the side edge and indicating position information of the side edge in the vehicle image, includes:
if a side edge is marked as present, a second loss value is calculated for an output item corresponding to the side edge and representing the position information of the side edge in the vehicle image by using a preset second loss function and a corresponding output value.
Optionally, the plurality of output items include output items indicating whether a left front edge, a left rear edge, a right front edge, and a right rear edge of the vehicle are present, respectively; and output items respectively representing position information of the left front edge, the left rear edge, the right front edge and the right rear edge in the vehicle image.
The embodiment of the invention also discloses a vehicle side edge determining device, which comprises:
the prediction content acquisition module is used for acquiring the vehicle image acquired by the camera and a pre-trained prediction model; the predictive model includes a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image;
a first output value acquisition module for acquiring output values of a plurality of output items output to the vehicle image by the prediction model;
a side edge existence determination module for determining whether a side edge exists based on the output value of the output item indicating whether the side edge of the vehicle exists;
and the side edge position determining module is used for determining the position information of the side edge in the vehicle image according to the output value of the output item corresponding to the determined side edge if the side edge is determined to exist.
Optionally, the side edge existence determining module includes:
the normalization processing submodule is used for performing normalization processing on the output values of the output items respectively representing whether the side edges of the vehicle exist or not;
an output value determination submodule for determining whether there is a normalized output value greater than a preset threshold;
and the side edge existence determining submodule is used for determining that the corresponding side edge exists if the normalized output value which is greater than the preset threshold exists.
Optionally, the method further comprises:
training the predictive model by:
a training content acquisition module for acquiring vehicle image samples and a prediction model, the prediction model comprising a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image;
a second output value obtaining module, configured to obtain output values of a plurality of output items output by the prediction model for the vehicle image sample;
a first loss value calculation module for calculating a first loss value for the output items respectively representing whether the side edges of the vehicle exist or not;
a second loss value calculation module for calculating a second loss value for the output items respectively representing the position information of the side edges in the vehicle image;
and the model training module is used for training the prediction model according to the first loss value and the second loss value.
Optionally, the first loss value calculation module includes:
and the first loss value operator module is used for calculating a first loss value by using a preset first loss function for the output items respectively representing whether the side edges of the vehicle exist or not.
Optionally, the vehicle image sample has annotation information indicating whether a side edge exists; the second loss value calculation module includes:
a first condition loss value calculation operator module, configured to calculate a second loss value according to a corresponding output value for an output item corresponding to a side edge and indicating position information of the side edge in the vehicle image if the side edge is marked as present;
and the second condition loss value operator module is used for determining that a second loss value is 0 for an output item which corresponds to the side edge and represents the position information of the side edge in the vehicle image if the side edge is marked as not present.
Optionally, the first case loss value operator module includes:
and a first condition loss value calculation unit, configured to calculate a second loss value using a preset second loss function and a corresponding output value for an output item corresponding to the side edge and indicating position information of the side edge in the vehicle image, if the side edge is marked as present.
Optionally, the plurality of output items include output items indicating whether a left front edge, a left rear edge, a right front edge, and a right rear edge of the vehicle are present, respectively; and output items respectively representing position information of the left front edge, the left rear edge, the right front edge and the right rear edge in the vehicle image.
The embodiment of the invention also discloses an electronic device, which comprises: a processor, a memory and a computer program stored on the memory and capable of running on the processor, which computer program, when being executed by the processor, carries out the steps of the vehicle lateral edge determination method as described above.
An embodiment of the present invention further discloses a computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, implements the steps of the vehicle side edge determining method as described above.
The embodiment of the invention has the following advantages:
in the embodiment of the invention, the vehicle image collected by the camera and the pre-trained prediction model can be obtained; the predictive model includes a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image; output values of a plurality of output items output by the prediction model to the vehicle image may be acquired; determining whether a side edge exists based on an output value of an output item indicating whether the side edge of the vehicle exists; and if the existence of the side edge is determined, determining the position information of the side edge in the vehicle image according to the output value of the output item corresponding to the side edge determined to exist. The embodiment of the invention can determine whether the side edge of the vehicle exists and the position information of the existing side edge according to the vehicle image acquired by the camera and the prediction model. Compared with the scheme of predicting the vehicle side edge through the laser radar, the embodiment of the invention can reduce the cost and improve the prediction accuracy.
Drawings
FIG. 1 is a flow chart of steps in a method for vehicle side edge determination according to an embodiment of the present invention;
FIG. 2 is a schematic representation of an image of a vehicle in an embodiment of the present invention;
FIG. 3 is a flowchart illustrating steps of a predictive model training method according to an embodiment of the present invention;
fig. 4 is a block diagram of a vehicle side edge determining apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Referring to fig. 1, a flowchart illustrating steps of a method for determining a vehicle side edge according to an embodiment of the present invention is shown, where the method may specifically include the following steps:
In an embodiment of the invention, the vehicle may be provided with a monocular camera by which images of the surrounding vehicle may be captured. In one example, the monocular camera may be located behind the vehicle's rearview mirror or in the front bumper.
The vehicle image is an image including a vehicle. Fig. 2 is a schematic diagram of a vehicle image according to an embodiment of the present invention. Assuming that the vehicles on the road are all rectangular parallelepipeds in shape, in the vehicle image, the vehicle may be modeled as a rectangular parallelepiped, and then the smallest rectangular parallelepiped that can surround the vehicle (without a rear view mirror) is selected as a surrounding frame in a three-dimensional space. The four vertical edges of the rectangular parallelepiped can be considered as the four side edges of the vehicle. The orientation of the side edges in the vehicle can be classified into four types of orientations: left posterior edge, left anterior edge, right posterior edge, right anterior edge.
As shown in table 1, it is a relation table between output items and corresponding semantic information in the embodiment of the present invention.
| Bit position | Semantic information |
| 1 | Presence or absence of left anterior arris |
| 2 | Position of left anterior edge in image |
| 3 | Whether or not the left rear edge exists |
| 4 | Position of left back edge in image |
| 5 | Presence or absence of right front edge |
| 6 | Position of right anterior edge in image |
| 7 | Whether or not the right rear edge exists |
| 8 | Position of right back edge in image |
TABLE 1
As shown in table one, the plurality of output items include output items respectively indicating whether a left front edge, a left rear edge, a right front edge and a right rear edge of the vehicle exist or not; and output items respectively representing position information of the left front edge, the left rear edge, the right front edge and the right rear edge in the vehicle image.
And 102, acquiring output values of a plurality of output items output by the prediction model to the vehicle image.
In the embodiment of the present invention, the output item indicating whether the side edge of the vehicle is present or not may be represented using a binary bit. The output item representing the position information of the side edge in the vehicle image may be represented using a floating-point number bit.
And 103, determining whether the side edge exists according to the output value of the output item indicating whether the side edge of the vehicle exists.
In an embodiment of the present invention, the step 103 may include the following sub-steps:
and a substep S11 of normalizing the output values of the output items each indicating the presence or absence of a side edge of the vehicle.
For the output values of the output items each indicating whether or not a side edge of the vehicle is present, the values may be normalized to [0-1] using a sigmoid function.
And a sub-step S12 of determining whether there is a normalized output value greater than a preset threshold.
And a substep S13, determining that a corresponding side edge exists if there is a normalized output value greater than a preset threshold.
If the normalized output value greater than the preset threshold exists, the side edge corresponding to the output item of the output value can be considered to exist. For example, the values are normalized to [0-1] using a sigmoid function for 1, 3, 5, 7 bit output values as shown in Table one. If the value of a bit is greater than 0.5, the bit may be considered to be present with respect to the side edge.
And 104, if the existence of the side edge is determined, determining the position information of the side edge in the vehicle image according to the output value of the output item corresponding to the determined existence of the side edge.
For example, as shown in table one, if the normalized values of the output values of 1, 3, and 5 bits are greater than the threshold, it is determined that the left front edge exists, the left rear edge exists, and the right front edge exists. The position of the front left edge in the image, the position of the rear left edge in the image and the position of the front right edge in the image can be determined according to the numerical values of 2, 4 and 6 bits respectively.
In the embodiment of the invention, the vehicle image collected by the camera and the pre-trained prediction model can be obtained; the predictive model includes a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image; output values of a plurality of output items output by the prediction model to the vehicle image may be acquired; determining whether a side edge exists based on an output value of an output item indicating whether the side edge of the vehicle exists; and if the existence of the side edge is determined, determining the position information of the side edge in the vehicle image according to the output value of the output item corresponding to the side edge determined to exist. The embodiment of the invention can determine whether the side edge of the vehicle exists and the position information of the existing side edge according to the vehicle image acquired by the camera and the prediction model. Compared with the scheme of predicting the vehicle side edge through the laser radar, the embodiment of the invention can reduce the cost and improve the prediction accuracy.
Referring to fig. 3, a flowchart illustrating steps of a predictive model training method according to an embodiment of the present invention is shown, where the method specifically includes the following steps:
The vehicle image samples may be pre-processed vehicle images used to train the predictive model, and the vehicle image samples may be captured by a camera.
In step 303, a first loss value is calculated for the output items respectively representing whether the side edges of the vehicle exist.
The loss values may be calculated using different calculation methods for the output items respectively indicating the presence or absence of a side edge of the vehicle and the output items respectively indicating the position information of the side edge in the vehicle image.
In this embodiment of the present invention, the step 303 may include: for the output items respectively representing whether the side edges of the vehicle exist or not, a first loss value is calculated by using a preset first loss function.
The output items each indicating the presence or absence of a side edge of the vehicle may be represented using binary bits. For binary represented output values, a first loss value may be calculated using a cross entropy function.
In the embodiment of the invention, the vehicle image sample is provided with marking information which represents whether a side edge exists or not; the marking information can be information for marking whether the side edge exists or not by a marking person manually.
The step 304 may include the following sub-steps:
if a side edge is marked as present, a sub-step S21 calculates a second loss value from a corresponding output value for an output item indicating position information of the side edge in the vehicle image, the output item corresponding to the side edge.
If the side edge is manually marked, calculating a second loss value according to the corresponding output value for the output item which represents the position information of the side edge in the vehicle image. In one example, the second loss value may be calculated using a preset second loss function and a corresponding output value.
In the embodiment of the present invention, the output item representing the position information of the side edge in the vehicle image may be represented using a floating-point number bit, and thus the second loss value may be calculated using the L2 loss function and the corresponding output value for the output item representing the position information of the side edge in the vehicle image.
In the sub-step S22, if a side edge is marked as not present, a second loss value of 0 is determined for an output item indicating position information of the side edge in the vehicle image corresponding to the side edge.
If the manual marking is that the side edge is not present, it makes no sense to calculate a loss value of the output item representing the position information of the side edge in the vehicle image, and therefore it can be determined that the second loss value is 0.
A total loss value may be calculated from the first loss value and the second loss value, and the total loss value may be transmitted back to the predictive model to update parameters of the predictive model. In one example, the first loss value may be multiplied by a corresponding weight to be added to the second loss value multiplied by the corresponding weight, and the added value may be used as the total loss value.
It should be noted that, for simplicity of description, the method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present invention is not limited by the illustrated order of acts, as some steps may occur in other orders or concurrently in accordance with the embodiments of the present invention. Further, those skilled in the art will appreciate that the embodiments described in the specification are presently preferred and that no particular act is required to implement the invention.
Referring to fig. 4, a block diagram of a vehicle side edge determining apparatus according to an embodiment of the present invention is shown, and may specifically include the following modules:
a prediction content obtaining module 401, configured to obtain a vehicle image acquired by a camera and a pre-trained prediction model; the predictive model includes a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image;
a first output value acquisition module 402 for acquiring output values of a plurality of output items output to the vehicle image by the prediction model;
a side edge existence determination module 403 for determining whether a side edge exists based on the output value of the output item indicating whether a side edge of the vehicle exists;
a side edge position determining module 404, configured to determine, if it is determined that a side edge exists, position information of the side edge in the vehicle image according to an output value of an output item corresponding to the determined existing side edge.
In an embodiment of the present invention, the side edge existence determining module 403 may include:
the normalization processing submodule is used for performing normalization processing on the output values of the output items respectively representing whether the side edges of the vehicle exist or not;
an output value determination submodule for determining whether there is a normalized output value greater than a preset threshold;
and the side edge existence determining submodule is used for determining that the corresponding side edge exists if the normalized output value which is greater than the preset threshold exists.
In an embodiment of the present invention, the prediction model may be trained by:
a training content acquisition module for acquiring vehicle image samples and a prediction model, the prediction model comprising a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image;
a second output value obtaining module, configured to obtain output values of a plurality of output items output by the prediction model for the vehicle image sample;
a first loss value calculation module for calculating a first loss value for the output items respectively representing whether the side edges of the vehicle exist or not;
a second loss value calculation module for calculating a second loss value for the output items respectively representing the position information of the side edges in the vehicle image;
and the model training module is used for training the prediction model according to the first loss value and the second loss value.
In an embodiment of the present invention, the first loss value calculating module may include:
and the first loss value operator module is used for calculating a first loss value by using a preset first loss function for the output items respectively representing whether the side edges of the vehicle exist or not.
In the embodiment of the invention, the vehicle image sample is provided with marking information which represents whether a side edge exists or not; the second loss value calculation module may include:
a first condition loss value calculation operator module, configured to calculate a second loss value according to a corresponding output value for an output item corresponding to a side edge and indicating position information of the side edge in the vehicle image if the side edge is marked as present;
and the second condition loss value operator module is used for determining that a second loss value is 0 for an output item which corresponds to the side edge and represents the position information of the side edge in the vehicle image if the side edge is marked as not present.
In this embodiment of the present invention, the first condition loss value operator module may include:
and a first condition loss value calculation unit, configured to calculate a second loss value using a preset second loss function and a corresponding output value for an output item corresponding to the side edge and indicating position information of the side edge in the vehicle image, if the side edge is marked as present.
In an embodiment of the present invention, the plurality of output items include output items indicating whether or not a left front edge, a left rear edge, a right front edge, and a right rear edge of the vehicle are present, respectively; and output items respectively representing position information of the left front edge, the left rear edge, the right front edge and the right rear edge in the vehicle image.
In the embodiment of the invention, the vehicle image collected by the camera and the pre-trained prediction model can be obtained; the predictive model includes a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image; output values of a plurality of output items output by the prediction model to the vehicle image may be acquired; determining whether a side edge exists based on an output value of an output item indicating whether the side edge of the vehicle exists; and if the existence of the side edge is determined, determining the position information of the side edge in the vehicle image according to the output value of the output item corresponding to the side edge determined to exist. The embodiment of the invention can determine whether the side edge of the vehicle exists and the position information of the existing side edge according to the vehicle image acquired by the camera and the prediction model. Compared with the scheme of predicting the vehicle side edge through the laser radar, the embodiment of the invention can reduce the cost and improve the prediction accuracy.
For the device embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, refer to the partial description of the method embodiment.
An embodiment of the present invention further provides an electronic device, including:
the vehicle side edge determining method comprises a processor, a memory and a computer program which is stored in the memory and can run on the processor, wherein when the computer program is executed by the processor, each process of the vehicle side edge determining method embodiment is realized, the same technical effect can be achieved, and in order to avoid repetition, the description is omitted.
The embodiment of the present invention further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the computer program implements each process of the embodiment of the vehicle side edge determining method, and can achieve the same technical effect, and in order to avoid repetition, details are not repeated here.
The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing terminal to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing terminal to cause a series of operational steps to be performed on the computer or other programmable terminal to produce a computer implemented process such that the instructions which execute on the computer or other programmable terminal provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications of these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the embodiments of the invention.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or terminal that comprises the element.
The present invention provides a vehicle side edge determining method and a vehicle side edge determining device, which are described in detail above, and the principle and the implementation of the present invention are explained herein by applying specific examples, and the description of the above examples is only used to help understand the method and the core idea of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. A vehicle side edge determining method, characterized by comprising:
acquiring a vehicle image acquired by a camera and a pre-trained prediction model; the predictive model includes a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image;
acquiring output values of a plurality of output items output by the prediction model to the vehicle image;
determining whether a side edge exists according to the output value of the output item indicating whether the side edge of the vehicle exists;
and if the existence of the side edge is determined, determining the position information of the side edge in the vehicle image according to the output value of the output item corresponding to the side edge determined to exist.
2. The method of claim 1, wherein the determining whether a side edge exists according to the output value of the output item corresponding to whether the side edge of the vehicle exists includes:
normalizing the output values of the output items respectively representing whether the side edges of the vehicle exist or not;
determining whether a normalized output value greater than a preset threshold value exists;
and if the normalized output value greater than the preset threshold value exists, determining that the corresponding side edge exists.
3. The method of claim 1, further comprising:
training the predictive model by:
obtaining a vehicle image sample and a predictive model, the predictive model comprising a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image;
obtaining output values of a plurality of output items output by the prediction model to the vehicle image sample;
calculating a first loss value for the output items respectively representing whether the side edges of the vehicle exist or not;
calculating a second loss value for the output items respectively representing the position information of the side edges in the vehicle image;
and training the prediction model according to the first loss value and the second loss value.
4. The method of claim 3, wherein said calculating a first loss value for said output terms each indicative of the presence or absence of a vehicle side edge comprises:
for the output items respectively representing whether the side edges of the vehicle exist or not, a first loss value is calculated by using a preset first loss function.
5. The method of claim 3, wherein the vehicle image sample has annotation information indicating whether a side edge is present; the calculating, for the output items respectively representing the position information of the side edges in the vehicle image, a second loss value including:
if the side edge is marked as existing, calculating a second loss value according to a corresponding output value for an output item corresponding to the side edge and representing the position information of the side edge in the vehicle image;
if a side edge is marked as not present, a second loss value of 0 is determined for the output item corresponding to the side edge and representing the position information of the side edge in the vehicle image.
6. The method of claim 5, wherein if a side edge is marked as present, calculating a second loss value from a corresponding output value for an output item corresponding to the side edge and representing position information of the side edge in the vehicle image comprises:
if a side edge is marked as present, a second loss value is calculated for an output item corresponding to the side edge and representing the position information of the side edge in the vehicle image by using a preset second loss function and a corresponding output value.
7. The method of claim 1, wherein the plurality of output items include output items that respectively indicate whether a left front edge, a left rear edge, a right front edge, and a right rear edge of a vehicle are present, and output items that respectively indicate position information of the left front edge, the left rear edge, the right front edge, and the right rear edge in the vehicle image.
8. A vehicle side edge determining apparatus, comprising:
the prediction content acquisition module is used for acquiring the vehicle image acquired by the camera and a pre-trained prediction model; the predictive model includes a plurality of output items; the plurality of output items include output items respectively indicating whether side edges of a vehicle exist, and output items respectively indicating position information of the side edges in the vehicle image;
a first output value acquisition module for acquiring output values of a plurality of output items output to the vehicle image by the prediction model;
a side edge existence determination module for determining whether a side edge exists based on the output value of the output item indicating whether the side edge of the vehicle exists;
and the side edge position determining module is used for determining the position information of the side edge in the vehicle image according to the output value of the output item corresponding to the determined side edge if the side edge is determined to exist.
9. An electronic device, comprising: processor, memory and computer program stored on the memory and executable on the processor, which computer program, when being executed by the processor, carries out the steps of the vehicle side edge determination method according to any one of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the vehicle side edge determination method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110004834.XA CN112784705A (en) | 2021-01-04 | 2021-01-04 | Vehicle side edge determining method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110004834.XA CN112784705A (en) | 2021-01-04 | 2021-01-04 | Vehicle side edge determining method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112784705A true CN112784705A (en) | 2021-05-11 |
Family
ID=75754011
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110004834.XA Pending CN112784705A (en) | 2021-01-04 | 2021-01-04 | Vehicle side edge determining method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112784705A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP4145392A1 (en) * | 2021-09-02 | 2023-03-08 | Guangzhou Xiaopeng Autopilot Technology Co., Ltd. | Method and apparatus for determining three-dimensional information of target object |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107844784A (en) * | 2017-12-08 | 2018-03-27 | 广东美的智能机器人有限公司 | Face identification method, device, computer equipment and readable storage medium storing program for executing |
| CN110276287A (en) * | 2019-06-17 | 2019-09-24 | 百度在线网络技术（北京）有限公司 | Method for detecting parking stalls, device, computer equipment and storage medium |
| CN112036389A (en) * | 2020-11-09 | 2020-12-04 | 天津天瞳威势电子科技有限公司 | Vehicle three-dimensional information detection method, device and equipment and readable storage medium |
-
2021
- 2021-01-04 CN CN202110004834.XA patent/CN112784705A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107844784A (en) * | 2017-12-08 | 2018-03-27 | 广东美的智能机器人有限公司 | Face identification method, device, computer equipment and readable storage medium storing program for executing |
| CN110276287A (en) * | 2019-06-17 | 2019-09-24 | 百度在线网络技术（北京）有限公司 | Method for detecting parking stalls, device, computer equipment and storage medium |
| CN112036389A (en) * | 2020-11-09 | 2020-12-04 | 天津天瞳威势电子科技有限公司 | Vehicle three-dimensional information detection method, device and equipment and readable storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 伊本贵士主编: "IoT最强教科书 完全版 5G时代物联网技术应用解密 人工智能（AI）的基石", 中国青年出版社, pages: 246 - 248 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP4145392A1 (en) * | 2021-09-02 | 2023-03-08 | Guangzhou Xiaopeng Autopilot Technology Co., Ltd. | Method and apparatus for determining three-dimensional information of target object |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10229332B2 (en) | Method and apparatus for recognizing obstacle of vehicle | |
| CN110263652B (en) | Laser point cloud data identification method and device | |
| CN107808390A (en) | Estimated using the object distance of the data from single camera | |
| US10937176B2 (en) | Object recognition apparatus | |
| DE102018100192A1 (en) | Generation of training data for automatic leak detection in vehicles | |
| CN110341621B (en) | Obstacle detection method and device | |
| JP2016194925A (en) | Method and device of detecting road boundary object | |
| US20150003669A1 (en) | 3d object shape and pose estimation and tracking method and apparatus | |
| US10913455B2 (en) | Method for the improved detection of objects by a driver assistance system | |
| EP3940487A1 (en) | Estimation of probability of collision with increasing severity level for autonomous vehicles | |
| CN113888568A (en) | Multilane optimization method, multilane optimization device, electronic apparatus, storage medium, and program | |
| CN113212442A (en) | Trajectory-aware vehicle driving analysis method and system | |
| CN112784705A (en) | Vehicle side edge determining method and device | |
| CN114919584A (en) | Motor vehicle fixed point target distance measuring method and device and computer readable storage medium | |
| EP3637311A1 (en) | Device and method for determining the altitude information of an object in an environment of a vehicle | |
| CN112560689B (en) | Parking space detection method and device, electronic equipment and storage medium | |
| CN112693450B (en) | Analysis method, analysis device, electronic equipment and analysis medium for optimizing automatic parking | |
| CN112926378B (en) | Vehicle side edge determining method and device | |
| CN112477868B (en) | Collision time calculation method and device, readable storage medium and computer equipment | |
| CN115140029A (en) | Safety capability detection method and device for automatic driving automobile | |
| US20220398837A1 (en) | Method for calculating a quality measure for assessing an object detection algorithm | |
| CN112733697B (en) | Method and device for determining yaw angle of vehicle | |
| US20230085898A1 (en) | Automatic cross-sensor calibration using object detections | |
| EP4336467A1 (en) | Method and apparatus for modeling object, storage medium, and vehicle control method | |
| CN116729364A (en) | Parking auxiliary method, parking auxiliary system, electronic equipment and vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |