CN110807123A - Vehicle length calculation method, device and system, computer equipment and storage medium - Google Patents
Vehicle length calculation method, device and system, computer equipment and storage medium Download PDFInfo
- Publication number
- CN110807123A CN110807123A CN201911034312.3A CN201911034312A CN110807123A CN 110807123 A CN110807123 A CN 110807123A CN 201911034312 A CN201911034312 A CN 201911034312A CN 110807123 A CN110807123 A CN 110807123A
- Authority
- CN
- China
- Prior art keywords
- data set
- vehicle
- vehicle length
- training
- picture
- 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
- 238000004364 calculation method Methods 0.000 title claims abstract description 47
- 238000012549 training Methods 0.000 claims abstract description 68
- 238000012360 testing method Methods 0.000 claims abstract description 67
- 238000000034 method Methods 0.000 claims description 22
- 238000004590 computer program Methods 0.000 claims description 14
- 238000010276 construction Methods 0.000 claims description 14
- 238000013095 identification testing Methods 0.000 claims description 8
- 230000003044 adaptive effect Effects 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 6
- 239000013589 supplement Substances 0.000 claims description 5
- 230000001502 supplementing effect Effects 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 abstract description 7
- 238000011161 development Methods 0.000 abstract description 5
- 238000002372 labelling Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 238000013527 convolutional neural network Methods 0.000 description 6
- 238000001514 detection method Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 230000018109 developmental process Effects 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000007689 inspection Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
-
- 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)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Library & Information Science (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Databases & Information Systems (AREA)
- Multimedia (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention provides a vehicle length calculation method, which comprises the following steps: constructing a picture data set; marking pictures in the data set to construct a marked data set; establishing a YOLO model, changing model parameters according to the format of the labeled data set, and training the training set and the labeled data set by using the model to obtain an initial weight file; constructing a vehicle length information table; calling an initial weight file, testing the test set and the labeled data set thereof by using the model, adjusting parameters according to the accuracy of results, and storing the test weight file; and identifying the picture to be detected by using the model, matching the identification result with the vehicle length information table, and acquiring the length information of the vehicle. The invention also provides a vehicle length calculating device, a vehicle length calculating system, computer equipment and a storage medium. The invention realizes the calculation of the length of the vehicle, is applied to the field of intelligent traffic, can promote the development of intelligent driving, can also be used for classifying the vehicle and provides basic data for traffic and logistics analysis.
Description
Technical Field
The invention relates to the field of intelligent transportation, in particular to a vehicle length calculation method, a vehicle length calculation device, a vehicle length calculation system, computer equipment and a storage medium.
Background
The target detection, also called target extraction, is an image segmentation based on target geometry and statistical characteristics, which combines the segmentation and identification of a target image into a whole, and the accuracy and real-time performance of the method are important capabilities of the whole system. Especially, in a complex scene, when a plurality of targets need to be processed in real time, automatic target extraction and identification are particularly important.
With the development of computer technology and the wide application of computer vision principle, the research of real-time tracking of targets by using computer image processing technology is more and more popular, and the dynamic real-time tracking and positioning of targets has wide application value in the aspects of intelligent traffic systems, intelligent monitoring systems, military target detection, positioning of surgical instruments in medical navigation operations and the like.
In the prior art, R-CNN (Convolutional Neural Networks, R-CNN (registers with CNN features), a method for applying large-scale Convolutional Neural Networks (CNNs) to candidate Regions from bottom to top to position and divide objects, and a YoLO (Real-Time Object Detection), which are used for Object Detection, are commonly used, but the two models can Only be used for classifying objects in pictures, can not be applied to the field of intelligent transportation, can not well classify the length of vehicles, and can not meet the development requirements of intelligent transportation systems.
Disclosure of Invention
The invention aims to provide a vehicle length calculation method, a vehicle length calculation device, a vehicle length calculation system, computer equipment and a storage medium, so that vehicle lengths can be classified to realize vehicle length calculation.
In order to achieve the above object, the present invention provides a vehicle length calculating method, including:
s201: constructing a picture data set, wherein the picture data set is divided into a training set and a testing set;
s202: calling a picture marking tool to mark the pictures in the picture data set to obtain corresponding marking results, and respectively storing the marking results and the pictures into different folders to construct a marked data set;
s203: establishing a YOLO model, changing parameters of the YOLO model according to the format of the labeled data set, and then training the training set and the labeled data set by using the YOLO model to obtain an initial weight file suitable for the current application environment;
s204: constructing a vehicle length information table according to the vehicle information on the network and the vehicle type of the current application environment;
s205: calling the initial weight file in the step S203, performing vehicle type identification test on the test set and the labeled data set thereof by using the YOLO model to obtain a test result, adjusting network hyper-parameters according to the accuracy of the test result to finally obtain an optimal solution, and storing a corresponding test weight file;
s206: and calling the test weight file in the step S205, identifying the type of the vehicle for the photo to be tested by using the YOLO model, matching the identification result with the vehicle length information table in the step S204, and acquiring the length information of the vehicle.
In another aspect, the present invention also provides a vehicle length calculating apparatus, including:
the data set construction module is used for constructing a picture data set and dividing the picture data set into a training set and a testing set;
the marking module is used for calling a picture marking tool to mark the picture in the picture data set to obtain a corresponding marking result, and storing the marking result and the picture into different folders respectively so as to construct a marked data set;
the training module is used for acquiring the format of the labeled data set, changing parameters of a YOLO model to enable the parameters to be adaptive to the format of the labeled data set, training the labeled data set by using the YOLO model, and acquiring a weight file suitable for the current application environment after the training is finished;
the vehicle length information building module is used for building a vehicle length information table according to the vehicle information on the network and the vehicle type of the current application environment;
the preprocessing module is configured to call the initial weight file in the step S203, perform a vehicle type identification test on the test set and the labeled data set thereof by using the YOLO model to obtain a test result, adjust a network hyper-parameter according to the accuracy of the test result, finally obtain an optimal solution, and store a corresponding test weight file; and
and the matching module is used for calling the test weight file, identifying the type of the vehicle for the photo to be tested by using the YOLO model, matching the identified result with the vehicle length information table and acquiring the length information of the vehicle.
Preferably, the data set building module comprises: the data set main body unit is set to acquire a field picture of the current application environment as a data set main body; the data set supplementing unit is used for acquiring pictures close to the scene pictures from a network as the supplement of the data set; and the data set dividing unit is used for dividing all pictures of the picture data set into a training set and a test set according to a preset proportion.
Preferably, the training module comprises: a label format acquiring unit configured to acquire a format of the label data set; the training unit is used for transmitting the format of the labeled data set to a GPU server, changing the parameters of a YOLO model by the GPU server to enable the parameters to be adaptive to the format of the labeled data set, and training the training set and the labeled data set; and a weight file acquiring unit configured to acquire a weight file suitable for the current application environment.
In another aspect, the present invention provides a vehicle length calculating system, including: the vehicle length calculation means according to the above; and the image acquisition device is connected with the vehicle length calculation device and is used for acquiring the field picture of the application scene.
In another aspect, the present invention also provides a computer device comprising a memory and a processor, the memory having stored therein a computer program which, when executed by the processor, causes the processor to implement the vehicle length calculation method according to the above.
In another aspect, the present invention also provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the vehicle length calculation method according to the above.
According to the vehicle length calculation method, the device, the computer equipment and the storage medium, the picture data set is constructed by using the picture of the current application scene, the YOLO model is used for training the data set to obtain the weight file suitable for the current application environment, the YOLO model calls the weight file to carry out vehicle marking and classification on the picture needing to be identified, the vehicle length information table is made, the marked vehicle is matched by using the information table, and the corresponding vehicle length information is obtained.
Drawings
FIG. 1 is a diagram of an application environment of a vehicle length calculation method according to an embodiment of the present invention;
FIG. 2 is a flow chart of a vehicle length calculation method according to an embodiment of the present invention;
fig. 3 is a block diagram of a vehicle length calculation apparatus according to an embodiment of the invention;
FIG. 4 is a block diagram of a vehicle length calculation system according to an embodiment of the present invention;
fig. 5 is a block diagram of the internal structure of a computer apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Fig. 1 is a diagram illustrating an application environment of a vehicle length calculating method according to an embodiment of the present invention, and as shown in fig. 1, the application environment includes a vehicle length calculating device 110, an image acquiring device 120, and a field vehicle 130.
In this embodiment, the vehicle length calculating device 110 may be integrated into a vehicle-mounted control system, and as a part of an intelligent driving system, calculates the length of a vehicle encountered during vehicle traveling, so as to avoid collision of the vehicle, and may also be fixedly disposed in a specific place, such as a parking lot, a toll gate, or a temporary transportation temporary inspection station, to analyze and calculate the passing vehicle, and of course, may also be used for collecting vehicle information, providing data support for system analysis, such as logistics analysis, and the like, which is not limited in this application. In consideration of real-time performance, accuracy and equipment complexity, the system is mainly suitable for parking lots, toll stations or temporary traffic temporary inspection stations and the like.
In this embodiment, the image acquiring device 120 may be a camera, or may be other devices with an image capturing function, such as a driving recorder, a mobile phone, a monitoring camera, and the like, and is connected to the vehicle length calculating device 110 in a wired or wireless manner.
In this embodiment, the on-site vehicle 130 may be any type of vehicle, and may be moving or stationary, the image of the on-site vehicle 130 acquired by the image acquisition device 120 may be a front surface or a side surface thereof, or at other angles, and further, the on-site vehicle 130 in the image acquired by the image acquisition device 120 may also be partially occluded. Thus, a vehicle image is acquired by the image acquiring means 120, and the image is calculated by the vehicle length calculating means 110, thereby obtaining the on-site vehicle length a.
The vehicle length calculation method provided by the invention can be applied to practical occasions, hardware equipment can be realized by completely utilizing the prior art, development and addition of new equipment are not needed, and the method is easy to realize.
The present invention provides a vehicle length calculating method 200, which is mainly illustrated in the present embodiment by the vehicle length calculating device 110 in fig. 1. As shown in fig. 2, the method may specifically include the following steps:
s201: constructing a picture data set, wherein the picture data set is divided into a training set and a testing set;
wherein the picture data set comprises a picture containing a vehicle.
The step S201 specifically includes:
s2011: acquiring a field picture of a current application environment as a data set main body;
s2012: acquiring pictures similar to the live pictures from a network as supplements of the data set;
s2013: and dividing all pictures of the picture data set into a training set and a testing set according to a preset ratio of 7: 3.
Wherein, the current application environment includes but is not limited to a road gate, a factory gate, etc., and the weighting coefficient is continuously updated and optimized in the subsequent process to adapt to the current application environment.
In the invention, the on-site picture of the application scene can be acquired by any equipment with an image acquisition function, such as a camera, and also can be other equipment with an image acquisition function, such as a vehicle data recorder, a mobile phone, a monitoring camera and the like.
In the invention, the way of acquiring the pictures on the network is mainly various automobile official networks and the automobile pictures in various specific application scenes, wherein the pictures from the specific application scenes are closer to the actual application occasions, and the identification accuracy of the vehicles in the actual application is improved.
In the invention, two basic functions of the data set are training and testing, the training is a basis, so that the YOLO model is suitable for practical occasions, the testing is a test on a training result and is a necessary link for putting into practice, and the problems of the model can be found through the testing, thereby further optimizing the model.
According to the invention, the construction data set is limited, so that the source of the data set is more fit with the actual application scene, and the accuracy of vehicle identification calculation is improved.
S202: calling a picture marking tool to mark the pictures in the picture data set to obtain corresponding marking results, storing the marking results and the pictures into different folders respectively, and ensuring that each picture has a corresponding marking result so as to construct a marking data set;
in the step S202, the image labeling tool is a conventional tool, such as MATLAB Training image label, and can be selected according to specific needs.
S203: establishing a YOLO model, changing parameters of the YOLO model according to the format of the labeled data set in the step S202 to enable the parameters to be adapted to the format of the labeled data set, then training the training set and the labeled data set thereof by using the YOLO model, and obtaining an initial weight file suitable for the current application environment after the training is finished.
Wherein the Yolo model is a Yolo v3 initial model obtained from an open source website. Since the YOLO model cannot be applied directly to the actual demand, the parameters need to be modified here.
The step S203 specifically includes:
s2031: acquiring the format of the labeled data set;
s2032: transmitting the format of the labeled data set to a GPU server, wherein the GPU server changes parameters of a YOLO model to enable the parameters to be adaptive to the format of the labeled data set, and trains the training set and the labeled data set thereof;
s2033: and acquiring a weight file which is obtained by the GPU server through training and is suitable for the current application environment.
In the invention, because a training process needs a large amount of calculation, the occupation of local operation resources is serious, and a GPU (graphic Processing Unit) provided by a server is used, so that a faster operation speed can be obtained. However, the prior art provides a plurality of cloud GPUs which can be directly used, and only local operation commands are changed and directly called. The method can reduce the occupation of local resources, obtain the calculation result more quickly, improve the working efficiency and facilitate the cost control.
S204: constructing a vehicle length information table according to the vehicle information on the network and the vehicle type of the current application environment;
the vehicle types are determined by vehicle information published on the network and vehicle information appearing in the picture data application scene, the number of the vehicle types is selected according to the training requirement in the step S203, for example, the lengths of a bus, a car and a tank car are calculated, when training is performed in the step S203, the three types of vehicle samples in the labeled data set are input into a YOLO model for training, corresponding three types of vehicle length information are input into the YOLO model, and after the training is finished, a new sample is input, the vehicle type is identified, and then the corresponding vehicle length information is returned. The vehicle types include, but are not limited to, vehicle types in the Chinese automobile classification standard.
In addition, the step S204 may be completed once, or may be updated according to the update timing of the network information, and the vehicle length information table does not need to be reconstructed every time the picture is identified.
S205: calling the initial weight file in the step S203, performing identification test on the vehicle type on the test set and the labeled data set thereof by using the YOLO model to obtain a test result comprising the vehicle type and the vehicle position, adjusting the network hyper-parameter according to the accuracy of the test result to finally obtain an optimal solution, and storing the corresponding test weight file. The network hyper-parameters include, but are not limited to, learning rate reduction strategy, iteration number, batch processing volume, batch number, iou threshold, etc.
Wherein, carry out the discernment test of vehicle type to a test set, include: and marking the vehicle position of the vehicle in the picture of the test set, and preliminarily judging the vehicle type of the vehicle.
Because the sizes of various vehicle types of the vehicles in China are uniform, the vehicle length corresponding to the type can be returned after the type of the vehicle is identified according to the YOLO model.
TABLE 1 vehicle type driver information Table
| Type of vehicle | Vehicle length |
| Mini-truck | 3.5m |
| Medium-sized cargo truck | 6m |
| Box type truck | 17.5m |
| Car with wheels | 4m |
| suv | 4.5m |
| Bus with movable front and rear wheels | 10m |
| Concrete tank truck | 10.5m |
S206: and calling the test weight file in the step S205, identifying the type of the vehicle for the photo to be tested by using the YOLO model, matching the identification result with the vehicle length information table in the step S204, and acquiring the length information of the vehicle.
In the invention, the steps S201 to S204 are only carried out during system construction, the system construction is completed, and the steps S205 to S206 are only required to be repeated during practical application. The vehicle detection and identification of the invention has better real-time performance than the prior art, 58 samples can be identified per second by using a 1080ti display card, and the vehicle length detection effect is achieved by using the potential inherent relationship between the vehicle type and the vehicle length information by using a template matching scheme, thereby avoiding the complexity of calculating the vehicle by arranging a plurality of cameras.
According to the vehicle length calculation method, the picture data set is constructed by using the pictures of the application scene of the method, the data set is trained by using the YOLO model to obtain the weight file suitable for the application environment, the YOLO model calls the weight file to perform vehicle marking on the pictures needing to be identified, a vehicle length information table is made, the marked vehicles are matched by using the information table, and corresponding vehicle length information is obtained. The method and the device can realize the calculation of the length of the vehicle for the vehicle picture acquired by the camera, are applied to the field of intelligent driving, can improve the recognition capability of an intelligent driving system for the vehicle, better plan the traveling route and prevent accidents, and can also be applied to the fields of traffic management, logistics analysis and the like.
As shown in fig. 3, a vehicle length calculating device 300 is implemented based on the vehicle length calculating method described above, and the structure of the vehicle length calculating device 300 is the same as that of the vehicle length calculating device 110 described above and shown in fig. 1. As shown in fig. 3, the vehicle length calculating device 300 includes:
a data set construction module 301 configured to construct a picture data set, and divide the picture data set into a training set and a test set;
the labeling module 302 is configured to call a picture labeling tool to label the picture in the picture data set to obtain a corresponding labeling result, and store the labeling result and the picture in different folders respectively, so as to construct a labeled data set;
a training module 303 configured to establish a YOLO model, change parameters of the YOLO model according to the format of the labeled data set, and then train the training set and the labeled data set thereof by using the YOLO model to obtain an initial weight file applicable to the current application environment;
the vehicle length information construction module 304 is configured to construct a vehicle length information table according to vehicle information on the network and the vehicle type of the current application environment;
the preprocessing module 305 is configured to call the initial weight file, perform a vehicle type identification test on the test set and the labeled data set thereof by using the YOLO model to obtain a test result, adjust a network hyper-parameter according to the accuracy of the test result, finally obtain an optimal solution, and store a corresponding test weight file; and
the matching module 306 is configured to call the test weight file, recognize the vehicle type of the photo to be tested by using the YOLO model, match the recognition result with the vehicle length information table, and acquire the length information of the vehicle.
In the present invention, the data set construction module 301, the labeling module 302, the training module 303, and the vehicle length information construction module 304 are only operated when a system is constructed, wherein the vehicle length information construction module 304 may be completed at one time when the system is constructed, or may be updated according to the update timing of network information, and it is not necessary to reconstruct the vehicle length information table every time image recognition is performed. The system is constructed, and the vehicle length can be analyzed and calculated only by the preprocessing module 305 and the matching module 306 in practical application.
In the invention, the Image marking tool is the prior art, such as MATLAB tracing Image Labeler and the like, and can be selected according to specific needs.
According to the vehicle length calculation device, a picture data set is constructed through pictures of an application scene, a YOLO model is used for training the data set to obtain a weight file suitable for the application environment, a vehicle length information table is manufactured, the YOLO model calls the weight file to mark vehicles on pictures needing to be identified, the marked vehicles are matched through the information table, and corresponding vehicle length information is obtained. Therefore, the method and the device can realize the calculation of the length of the vehicle for the vehicle picture acquired by the camera, are applied to the field of intelligent driving, can improve the recognition capability of an intelligent driving system for the vehicle, better plan the traveling route, prevent accidents, and can also be applied to the fields of traffic management, logistics analysis and the like.
Wherein the data set construction module 301 comprises:
the data set main body unit is set to acquire a field picture of the current application environment as a data set main body;
the data set supplementing unit is used for acquiring pictures close to the scene pictures from a network as the supplement of the data set; and
and the data set dividing unit is set to divide all pictures of the picture data set into a training set and a test set according to a preset proportion.
In the invention, the way for the data set supplementing unit to acquire the pictures is mainly various automobile official networks and automobile pictures in various specific application scenes, wherein the pictures from the specific application scenes are closer to the actual application occasions, and the identification accuracy of the vehicles in the actual application is improved.
In the invention, the data set dividing unit is used for dividing the data set, two basic functions of the data set are training and testing, the training is a basis, so that the YOLO model is suitable for practical occasions, the testing is a test on training results and is a necessary link for putting into practice, and the problems of the model can be found through the testing, thereby further optimizing the model.
The invention limits the construction data set, so that the source of the data set is more fit with the actual application scene, thereby improving the accuracy of vehicle identification and calculation.
The training module 303 includes:
a label format acquiring unit configured to acquire a format of the label data set;
the training unit is used for transmitting the format of the labeled data set to a GPU server, changing the parameters of a YOLO model by the GPU server to enable the parameters to be adaptive to the format of the labeled data set, and training the training set and the labeled data set; and
and the weight file acquisition unit is used for acquiring the weight file which is obtained by the GPU server through training and is suitable for the current application environment.
In the invention, because a training process needs a large amount of calculation, the occupation of local operation resources is serious, and a GPU (graphic Processing Unit) provided by a server is used, so that a faster operation speed can be obtained. However, the prior art provides a plurality of cloud GPUs which can be directly used, and only local operation commands are changed and directly called. The method can reduce the occupation of local resources, obtain the calculation result more quickly, improve the working efficiency and facilitate the cost control.
According to another embodiment of the present invention, there is also provided a vehicle length calculation system 400, as shown in fig. 4, the system 400 including:
a vehicle length calculating means 410, the vehicle length calculating means 410 being identical in construction to the vehicle length calculating means 110, 300 described above; and
and an image acquiring device 420 connected to the vehicle length calculating device, wherein the image acquiring device 420 has the same structure as the image acquiring device 120 described above and is used for acquiring a live picture of an application scene.
In this embodiment, the vehicle length calculating device 410 may be integrated in a vehicle-mounted control system, and as a part of an intelligent driving system, calculates the length of a vehicle encountered during vehicle traveling, so as to avoid collision of the vehicle, and may also be fixedly disposed in a specific place, such as a parking lot, a toll gate, or a temporary transportation temporary inspection station, to analyze and calculate the passing vehicle, and of course, may also be used for collecting vehicle information, and providing data support for system analysis, such as logistics analysis, and the like, which is not limited in this application.
In this embodiment, the image acquiring device 420 may be a camera, or may be other devices with an image capturing function, such as a driving recorder, a mobile phone, a monitoring camera, and the like, and is connected to the vehicle length calculating device 110 in a wired or wireless manner.
The vehicle length calculation system 400 provided by the embodiment of the invention is applied to practical occasions, hardware equipment can be realized by fully utilizing the prior art, development and addition of new equipment are not needed, and the implementation is easy.
Fig. 5 shows an internal structural diagram of the computer device 500 in one embodiment. The computer device 500 may specifically be the vehicle length calculating means 110 as shown in fig. 1. As shown in fig. 5, the computer apparatus 500 includes a processor, a memory, a network interface, an input device, and a display screen connected through a system bus. Wherein the memory includes a non-volatile storage medium and a memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program that, when invoked by a processor, causes the processor to perform the vehicle length calculation method 200 described above. The memory is configured to read the computer program from the non-volatile storage medium storage such that the processor invokes the computer program and associated files from the memory.
The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the computer equipment, an external keyboard, a touch pad or a mouse and the like.
Those skilled in the art will appreciate that the architecture shown in fig. 5 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a vehicle length calculating apparatus provided herein may be implemented in the form of a computer program that is executable on a computer device such as that shown in fig. 5. The memory of the computer device may store therein various program modules constituting the vehicle length calculating means, such as a data set constructing module 301, a labeling module 302, a training module 303, a vehicle length information constructing module 304, a preprocessing module 305, and a matching module 306 shown in fig. 3. The respective program modules constitute computer programs that cause the processor to execute the steps in the vehicle length calculation method of the respective embodiments of the present application described in the present specification.
For example, the computer device shown in fig. 5 may execute the step S201 in the vehicle length calculating method 200 described above through the data set constructing module 301 in one vehicle length calculating apparatus 300 shown in fig. 3, the computer device may execute the step S202 through the labeling module 302, the computer device may execute the step S203 through the training module 303, the computer device may execute the step S204 through the vehicle length information constructing module 304, the computer device may execute the step S205 through the preprocessing module 305, and the computer device may execute the step S20205 through the matching module 306.
In one embodiment, a computer device is proposed, the computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the vehicle length calculation method 200 described above when executing the computer program, i.e. performing the steps of:
constructing a picture data set, wherein the picture data set is divided into a training set and a testing set;
calling a picture marking tool to mark the pictures in the picture data set to obtain corresponding marking results, and respectively storing the marking results and the pictures into different folders to construct a marked data set;
establishing a YOLO model, changing parameters of the YOLO model according to the format of the labeled data set, and then training the training set and the labeled data set by using the YOLO model to obtain an initial weight file suitable for the current application environment;
constructing a vehicle length information table according to the vehicle information on the network and the vehicle type of the current application environment;
calling the initial weight file, performing vehicle type identification test on the test set and the labeled data set thereof by using the YOLO model to obtain a test result, adjusting network hyper-parameters according to the accuracy of the test result to finally obtain an optimal solution, and storing a corresponding test weight file;
and calling the test weight file, identifying the type of the vehicle for the photo to be tested by using the YOLO model, matching the identification result with the vehicle length information table in the step S204, and acquiring the length information of the vehicle.
In one embodiment, a computer readable storage medium is provided, having stored thereon a computer program which, when executed by a processor, causes the processor to implement the vehicle length calculation method 200 described above, i.e. to perform the steps of:
constructing a picture data set, wherein the picture data set is divided into a training set and a testing set;
calling a picture marking tool to mark the pictures in the picture data set to obtain corresponding marking results, and respectively storing the marking results and the pictures into different folders to construct a marked data set;
establishing a YOLO model, changing parameters of the YOLO model according to the format of the labeled data set, and then training the training set and the labeled data set by using the YOLO model to obtain an initial weight file suitable for the current application environment;
constructing a vehicle length information table according to the vehicle information on the network and the vehicle type of the current application environment;
calling the initial weight file, performing vehicle type identification test on the test set and the labeled data set thereof by using the YOLO model to obtain a test result, adjusting network hyper-parameters according to the accuracy of the test result to finally obtain an optimal solution, and storing a corresponding test weight file;
and calling the test weight file, identifying the type of the vehicle for the photo to be tested by using the YOLO model, matching the identification result with the vehicle length information table in the step S204, and acquiring the length information of the vehicle.
It should be understood that, although the steps in the flowcharts of the embodiments of the present invention are shown in sequence as indicated by the arrows, the steps are not necessarily performed in sequence as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a portion of the steps in various embodiments may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a non-volatile computer-readable storage medium, and can include the processes of the embodiments of the methods described above when the program is executed. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the embodiments described above may be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the embodiments described above are not described, but should be considered as being within the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (9)
1. A vehicle length calculation method, characterized by comprising:
s201: constructing a picture data set, wherein the picture data set is divided into a training set and a testing set;
s202: calling a picture marking tool to mark the pictures in the picture data set to obtain corresponding marking results, and respectively storing the marking results and the pictures into different folders to construct a marked data set;
s203: establishing a YOLO model, changing parameters of the YOLO model according to the format of the labeled data set, and then training the training set and the labeled data set by using the YOLO model to obtain an initial weight file suitable for the current application environment;
s204: constructing a vehicle length information table according to the vehicle information on the network and the vehicle type of the current application environment;
s205: calling the initial weight file in the step S203, performing vehicle type identification test on the test set and the labeled data set thereof by using the YOLO model to obtain a test result, adjusting network hyper-parameters according to the accuracy of the test result to finally obtain an optimal solution, and storing a corresponding test weight file;
s206: and calling the test weight file in the step S205, identifying the type of the vehicle for the photo to be tested by using the YOLO model, matching the identification result with the vehicle length information table in the step S204, and acquiring the length information of the vehicle.
2. The vehicle length calculation method according to claim 1, wherein the step S201 includes:
s2011: acquiring a field picture of a current application environment as a data set main body;
s2012: acquiring pictures similar to the live pictures from a network as supplements of the data set;
s2013: and dividing all pictures of the picture data set into a training set and a testing set according to a preset proportion.
3. The vehicle length calculating method according to claim 1, wherein the step S203 includes:
s2031: acquiring the format of the labeled data set;
s2032: transmitting the format of the labeled data set to a GPU server, wherein the GPU server changes parameters of a YOLO model to enable the parameters to be adaptive to the format of the labeled data set, and trains the training set and the labeled data set thereof;
s2033: and acquiring a weight file suitable for the current application environment.
4. A vehicle length calculation apparatus, characterized in that the apparatus comprises:
the data set construction module is used for constructing a picture data set and dividing the picture data set into a training set and a testing set;
the marking module is used for calling a picture marking tool to mark the picture in the picture data set to obtain a corresponding marking result, and storing the marking result and the picture into different folders respectively so as to construct a marked data set;
the training module is used for acquiring the format of the labeled data set, changing parameters of a YOLO model to enable the parameters to be adaptive to the format of the labeled data set, training the labeled data set by using the YOLO model, and acquiring a weight file suitable for the current application environment after the training is finished;
the vehicle length information building module is used for building a vehicle length information table according to the vehicle information on the network and the vehicle type of the current application environment;
the preprocessing module is configured to call the initial weight file in the step S203, perform a vehicle type identification test on the test set and the labeled data set thereof by using the YOLO model to obtain a test result, adjust a network hyper-parameter according to the accuracy of the test result, finally obtain an optimal solution, and store a corresponding test weight file; and
and the matching module is used for calling the test weight file, identifying the type of the vehicle for the photo to be tested by using the YOLO model, matching the identified result with the vehicle length information table and acquiring the length information of the vehicle.
5. The vehicle length calculation device of claim 4, wherein the data set construction module comprises:
the data set main body unit is set to acquire a field picture of the current application environment as a data set main body;
the data set supplementing unit is used for acquiring pictures close to the scene pictures from a network as the supplement of the data set; and
and the data set dividing unit is set to divide all pictures of the picture data set into a training set and a test set according to a preset proportion.
6. The vehicle length calculation device of claim 4, wherein the training module comprises:
a label format acquiring unit configured to acquire a format of the label data set;
the training unit is used for transmitting the format of the labeled data set to a GPU server, changing the parameters of a YOLO model by the GPU server to enable the parameters to be adaptive to the format of the labeled data set, and training the training set and the labeled data set; and
a weight file obtaining unit configured to obtain a weight file suitable for a current application environment.
7. A vehicle length calculation system, the system comprising:
the vehicle length calculation device according to one of claims 4 to 6; and
and the image acquisition device is connected with the vehicle length calculation device and is used for acquiring the field picture of the application scene.
8. A computer arrangement, characterized by comprising a memory and a processor, the memory having stored therein a computer program which, when executed by the processor, causes the processor to carry out a vehicle length calculation method according to one of claims 1-3.
9. A computer-readable storage medium, characterized in that a computer program is stored thereon, which computer program, when being executed by a processor, carries out a vehicle length calculation method according to one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911034312.3A CN110807123A (en) | 2019-10-29 | 2019-10-29 | Vehicle length calculation method, device and system, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911034312.3A CN110807123A (en) | 2019-10-29 | 2019-10-29 | Vehicle length calculation method, device and system, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110807123A true CN110807123A (en) | 2020-02-18 |
Family
ID=69489400
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911034312.3A Pending CN110807123A (en) | 2019-10-29 | 2019-10-29 | Vehicle length calculation method, device and system, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110807123A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111401162A (en) * | 2020-03-05 | 2020-07-10 | 上海眼控科技股份有限公司 | Illegal auditing method for muck vehicle, electronic device, computer equipment and storage medium |
| CN111598158A (en) * | 2020-05-14 | 2020-08-28 | 汇鼎数据科技(上海)有限公司 | Storage inventory state sensing method based on AI analysis technology |
| CN111598154A (en) * | 2020-05-14 | 2020-08-28 | 汇鼎数据科技(上海)有限公司 | AI analysis technology-based vehicle identification perception method |
| CN111626106A (en) * | 2020-04-17 | 2020-09-04 | 惠州市德赛西威智能交通技术研究院有限公司 | Camera vehicle detection rate statistical method and device |
| CN113450573A (en) * | 2020-03-25 | 2021-09-28 | 重庆翼动科技有限公司 | Traffic monitoring method and traffic monitoring system based on unmanned aerial vehicle image recognition |
| CN113850929A (en) * | 2021-09-18 | 2021-12-28 | 广州文远知行科技有限公司 | Display method, device, equipment and medium for processing marked data stream |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018157862A1 (en) * | 2017-03-02 | 2018-09-07 | 腾讯科技（深圳）有限公司 | Vehicle type recognition method and device, storage medium and electronic device |
| CN108875600A (en) * | 2018-05-31 | 2018-11-23 | 银江股份有限公司 | A kind of information of vehicles detection and tracking method, apparatus and computer storage medium based on YOLO |
| CN109472301A (en) * | 2018-10-26 | 2019-03-15 | 上海新增鼎数据科技有限公司 | A kind of Vehicle length calculation method, device, system and computer equipment |
-
2019
- 2019-10-29 CN CN201911034312.3A patent/CN110807123A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018157862A1 (en) * | 2017-03-02 | 2018-09-07 | 腾讯科技（深圳）有限公司 | Vehicle type recognition method and device, storage medium and electronic device |
| CN108875600A (en) * | 2018-05-31 | 2018-11-23 | 银江股份有限公司 | A kind of information of vehicles detection and tracking method, apparatus and computer storage medium based on YOLO |
| CN109472301A (en) * | 2018-10-26 | 2019-03-15 | 上海新增鼎数据科技有限公司 | A kind of Vehicle length calculation method, device, system and computer equipment |
Non-Patent Citations (1)
| Title |
|---|
| 余胜;陈敬东;王新余;: "基于深度学习的复杂场景下车辆识别方法" * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111401162A (en) * | 2020-03-05 | 2020-07-10 | 上海眼控科技股份有限公司 | Illegal auditing method for muck vehicle, electronic device, computer equipment and storage medium |
| CN113450573A (en) * | 2020-03-25 | 2021-09-28 | 重庆翼动科技有限公司 | Traffic monitoring method and traffic monitoring system based on unmanned aerial vehicle image recognition |
| CN111626106A (en) * | 2020-04-17 | 2020-09-04 | 惠州市德赛西威智能交通技术研究院有限公司 | Camera vehicle detection rate statistical method and device |
| CN111598158A (en) * | 2020-05-14 | 2020-08-28 | 汇鼎数据科技(上海)有限公司 | Storage inventory state sensing method based on AI analysis technology |
| CN111598154A (en) * | 2020-05-14 | 2020-08-28 | 汇鼎数据科技(上海)有限公司 | AI analysis technology-based vehicle identification perception method |
| CN111598158B (en) * | 2020-05-14 | 2023-05-12 | 汇鼎数据科技(上海)有限公司 | Warehouse inventory state sensing method based on AI analysis technology |
| CN113850929A (en) * | 2021-09-18 | 2021-12-28 | 广州文远知行科技有限公司 | Display method, device, equipment and medium for processing marked data stream |
| CN113850929B (en) * | 2021-09-18 | 2023-05-26 | 广州文远知行科技有限公司 | Display method, device, equipment and medium for processing annotation data stream |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110807123A (en) | Vehicle length calculation method, device and system, computer equipment and storage medium | |
| DE102011117585B4 (en) | Systems and methods for tracking objects | |
| CN105354568A (en) | Convolutional neural network based vehicle logo identification method | |
| CN109711264B (en) | Method and device for detecting occupation of bus lane | |
| CN111814623A (en) | Vehicle lane departure visual detection method based on deep neural network | |
| CN110991272A (en) | Multi-target vehicle track identification method based on video tracking | |
| Mahaur et al. | Road object detection: a comparative study of deep learning-based algorithms | |
| CN111582339B (en) | Vehicle detection and recognition method based on deep learning | |
| CN114565895B (en) | Security monitoring system and method based on intelligent society | |
| CN112613344B (en) | Vehicle track occupation detection method, device, computer equipment and readable storage medium | |
| DE102019113389A1 (en) | SYSTEMS AND METHODS FOR A PREDICTIVE ENTITY BEHAVIOR | |
| CN108241829A (en) | Vehicle travels image-recognizing method | |
| CN115018016A (en) | Method and system for identifying lane changing intention of manually-driven vehicle | |
| CN115546519B (en) | Matching method of image and millimeter wave radar target for extracting pseudo-image features | |
| CN113780200A (en) | Computer vision-based pavement multi-disease area detection and positioning method | |
| CN115830399A (en) | Classification model training method, apparatus, device, storage medium, and program product | |
| CN111753592A (en) | Traffic sign recognition method, traffic sign recognition device, computer equipment and storage medium | |
| CN112613434A (en) | Road target detection method, device and storage medium | |
| CN116824543A (en) | Automatic driving target detection method based on OD-YOLO | |
| Uzar et al. | Performance analysis of YOLO versions for automatic vehicle detection from UAV images | |
| CN110751040A (en) | Three-dimensional object detection method and device, electronic equipment and storage medium | |
| CN105335758A (en) | Model identification method based on video Fisher vector descriptors | |
| CN110097077B (en) | Point cloud data classification method and device, computer equipment and storage medium | |
| CN115830555A (en) | Target identification method based on radar point cloud, storage medium and equipment | |
| CN115376129A (en) | Mining area pavement semantic segmentation method, device, chip, terminal and computer equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200218 |
|
| WD01 | Invention patent application deemed withdrawn after publication |