CN110555347B - Vehicle target identification method and device with dangerous cargo-carrying behavior and electronic equipment - Google Patents
Vehicle target identification method and device with dangerous cargo-carrying behavior and electronic equipment Download PDFInfo
- Publication number
- CN110555347B CN110555347B CN201810557515.XA CN201810557515A CN110555347B CN 110555347 B CN110555347 B CN 110555347B CN 201810557515 A CN201810557515 A CN 201810557515A CN 110555347 B CN110555347 B CN 110555347B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- area
- image
- cargo
- vehicle target
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/29—Graphical models, e.g. Bayesian networks
-
- 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)
- 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)
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The embodiment of the invention provides a vehicle target identification method and device with dangerous cargo carrying behaviors and electronic equipment, wherein the vehicle target identification method with the dangerous cargo carrying behaviors comprises the following steps: carrying out vehicle target detection on the image to be detected, and determining the area of the vehicle target in the image to be detected; adjusting the area where the vehicle target is located according to the position relation of the area where the vehicle target is located in the image to be detected, and positioning a vehicle cargo-carrying interested area in the image to be detected, wherein the area where the vehicle target is located is an area which only contains the head or the tail of the vehicle target in the image to be detected, and the vehicle cargo-carrying interested area comprises cargo-carrying characteristics of the vehicle target; and classifying and identifying the vehicle targets in the vehicle freight carrying interested region by adopting a pre-trained neural network model, and judging whether the vehicle targets are the vehicle targets with dangerous freight carrying behaviors. The vehicle target with dangerous cargo carrying behavior can be accurately identified through the scheme.
Description
Technical Field
The invention relates to the technical field of intelligent transportation, in particular to a vehicle target identification method and device with dangerous cargo carrying behaviors and electronic equipment.
Background
In public transportation, dangerous cargo carrying behaviors of vehicles, such as overloading cargos in trucks shown in fig. 1a and 1b, overloading cargos in tricycles shown in fig. 1c and 1d, overloading cargos in car roofs shown in fig. 1e and 1f, binding cargos and the like, are frequently generated, and the dangerous cargo carrying behaviors of the vehicles easily cause traffic safety accidents. Therefore, in order to facilitate management work of public transportation safety-related departments, there is a need to identify such vehicles on roads that have dangerous cargo-carrying behavior. In the traditional identification method, vehicles with dangerous cargo carrying behaviors are screened out from a large amount of video monitoring image data in a manual screening mode, the data volume of the video monitoring image data is often very large, a large amount of manual time and energy are consumed, and the situations of missed detection and false detection are easy to occur.
In recent years, with the rapid development of artificial intelligence, deep learning methods are becoming mainstream techniques for target recognition in images. In public transportation, accurate identification of whether a vehicle has dangerous cargo-carrying behavior in an identification method based on deep learning cannot be carried out.
Disclosure of Invention
The embodiment of the invention aims to provide a method and a device for identifying a vehicle target with dangerous cargo-carrying behaviors and electronic equipment, so as to accurately identify the vehicle target with the dangerous cargo-carrying behaviors. The specific technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides a method for identifying a vehicle target with dangerous cargo-carrying behavior, where the method includes:
carrying out vehicle target detection on an image to be detected, and determining the area of a vehicle target in the image to be detected;
according to the position relation of the area where the vehicle target is located in the image to be detected, adjusting the area where the vehicle target is located, and positioning a vehicle loading interested area in the image to be detected, wherein the area where the vehicle target is located is an area which only contains the head or the tail of the vehicle target in the image to be detected, and the vehicle loading interested area comprises loading characteristics of the vehicle target;
and classifying and identifying the vehicle targets in the vehicle freight carrying interested region by adopting a pre-trained neural network model, and judging whether the vehicle targets are vehicle targets with dangerous freight carrying behaviors.
Optionally, before the adjusting the region where the vehicle target is located according to the position relationship of the region where the vehicle target is located in the image to be detected, and locating the region of interest for vehicle cargo in the image to be detected, the method further includes:
judging whether the size of the area where the vehicle target is located is larger than a preset size or not;
if so, adjusting the area of the vehicle target according to the position relation of the area of the vehicle target in the image to be detected, and positioning the region of interest for vehicle cargo in the image to be detected.
Optionally, the adjusting the region where the vehicle target is located according to the position relationship of the region where the vehicle target is located in the image to be detected, and locating the region of interest for vehicle cargo in the image to be detected includes:
determining a preset coordinate point in the area where the vehicle target is located;
and amplifying the region where the vehicle target is located according to a preset proportion by taking the preset coordinate point as a reference to obtain a vehicle cargo region of interest.
Optionally, the adjusting the region where the vehicle target is located according to the position relationship of the region where the vehicle target is located in the image to be detected, and locating the region of interest for vehicle cargo in the image to be detected includes:
Acquiring the width of a lower border of an area where the vehicle target is located, wherein the area where the vehicle target is located is a rectangular area where the head or the tail of the vehicle target is located;
the vehicle cargo area of interest is located by setting a midpoint of the lower frame to a lower frame midpoint of the vehicle cargo area of interest, setting a first height to a height of the vehicle cargo area of interest, setting a first width to a width of the vehicle cargo area of interest, wherein the first height is a product of the width of the lower frame and a first preset multiple, and the first width is a product of the width of the lower frame and a second preset multiple.
Optionally, the training process of the neural network model includes:
obtaining a plurality of positive samples containing vehicle objects having dangerous cargo carrying behavior and a plurality of negative samples containing vehicle objects not having dangerous cargo carrying behavior;
extracting each positive sample image and each negative sample image according to the coordinate information of the vehicle loading interested area containing the vehicle target in each calibrated positive sample and each calibrated negative sample, wherein the positive sample image is an image in the vehicle loading interested area in the positive sample, and the negative sample image is an image in the vehicle loading interested area in the negative sample;
And training the neural network model based on each positive sample image and each negative sample image.
In a second aspect, embodiments of the present invention provide a vehicle object recognition apparatus having dangerous cargo-carrying behavior, the apparatus comprising:
the detection module is used for detecting a vehicle target in an image to be detected and determining the area of the vehicle target in the image to be detected;
the positioning module is used for adjusting the area where the vehicle target is located according to the position relation of the area where the vehicle target is located in the image to be detected, and positioning a vehicle loading interested area in the image to be detected, wherein the area where the vehicle target is located is an area which only contains the head or the tail of the vehicle target in the image to be detected, and the vehicle loading interested area comprises loading characteristics of the vehicle target;
and the recognition module is used for classifying and recognizing the vehicle targets in the vehicle cargo-carrying interested region by adopting a pre-trained neural network model and judging whether the vehicle targets are the vehicle targets with dangerous cargo-carrying behaviors or not.
Optionally, the apparatus further comprises:
the judging module is used for judging whether the size of the area where the vehicle target is located is larger than a preset size or not;
And the positioning module is specifically used for adjusting the area where the vehicle target is located according to the position relation of the area where the vehicle target is located in the image to be detected if the judgment result of the judgment module is yes, and positioning the region of interest for vehicle cargo in the image to be detected.
Optionally, the positioning module is specifically configured to:
determining a preset coordinate point in the area where the vehicle target is located;
and amplifying the region where the vehicle target is located according to a preset proportion by taking the preset coordinate point as a reference to obtain a vehicle cargo region of interest.
Optionally, the positioning module is specifically configured to:
acquiring the width of a lower frame of an area where the vehicle target is located, wherein the area where the vehicle target is located is a rectangular area where the head or the tail of the vehicle target is located;
the vehicle cargo area of interest is located by setting a midpoint of the lower frame to a lower frame midpoint of the vehicle cargo area of interest, setting a first height to a height of the vehicle cargo area of interest, and setting a first width to a width of the vehicle cargo area of interest, wherein the first height is a product of a width of the lower frame and a first preset multiple, and the first width is a product of a width of the lower frame and a second preset multiple.
Optionally, the apparatus further comprises:
an acquisition module for acquiring a plurality of positive samples containing vehicle objects having dangerous cargo carrying behavior and a plurality of negative samples containing vehicle objects not having dangerous cargo carrying behavior;
the extraction module is used for extracting each positive sample image and each negative sample image according to the coordinate information of the vehicle loading interested area containing the vehicle target in each calibrated positive sample and each calibrated negative sample, wherein the positive sample image is an image in the vehicle loading interested area in the positive sample, and the negative sample image is an image in the vehicle loading interested area in the negative sample;
and the training module is used for training the neural network model based on each positive sample image and each negative sample image.
In a third aspect, embodiments of the present invention provide an electronic device, including a processor and a memory, wherein,
the memory is used for storing a computer program;
the processor is configured to implement the method steps of the first aspect of the embodiment of the present invention when executing the computer program stored in the memory.
According to the vehicle target identification method, the device and the electronic equipment with the dangerous cargo-carrying behavior, the region where the vehicle target is located in the image to be detected is obtained by detecting the vehicle target in the image to be detected, the region where the vehicle target is located is adjusted according to the position relation of the region where the vehicle target is located in the image to be detected, the region of interest containing cargo-carrying characteristics of the vehicle is located, and the vehicle target in the region of interest containing cargo-carrying characteristics of the vehicle is classified and identified by adopting a pre-trained neural network model, so that whether the vehicle target is the vehicle target with the dangerous cargo-carrying behavior can be judged. Compared with a target without dangerous cargo-carrying behaviors, the vehicle target with dangerous cargo-carrying behaviors has different peripheral outline sizes, and the region where the vehicle target is located is the region only containing the head or the tail of the vehicle target, so that the region where the vehicle target is located is adjusted to ensure that the adjusted region of interest for vehicle cargo-carrying can include the cargo-carrying characteristics of the vehicle target, and then a pre-trained neural network model is adopted to accurately classify and identify whether the vehicle target has dangerous cargo-carrying behaviors.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1a is a schematic view of a truck overloaded with cargo;
FIG. 1b is a schematic view of another truck overloaded with cargo;
FIG. 1c is a schematic view of a tricycle overloaded with cargo;
FIG. 1d is a schematic view of another tricycle overloading cargo;
FIG. 1e is a schematic view of a sedan top mounted bicycle;
FIG. 1f is a schematic view of another sedan top mount bicycle;
FIG. 2 is a simplified flow diagram of a method for identifying a vehicle target having dangerous cargo carrying behavior according to an embodiment of the present invention;
FIG. 3 is a flowchart illustrating a method for identifying objects of a vehicle with dangerous cargo-carrying behavior according to an embodiment of the present invention;
FIG. 4 is a flowchart illustrating a method for locating a cargo region of interest of a vehicle according to an embodiment of the present invention;
FIG. 5a is a schematic view of a vehicle loading area of interest based on a vehicle head according to an embodiment of the present invention;
FIG. 5b is a schematic illustration of a vehicle cargo area of interest based on the rear of a vehicle in accordance with an embodiment of the present invention;
FIG. 6 is a schematic flow chart diagram of a vehicle cargo area of interest locating method in accordance with another embodiment of the present invention;
FIG. 7a is a diagram illustrating the positioning effect of the area of interest for loading the vehicle based on the vehicle head according to the embodiment of the invention;
FIG. 7b is a graphical illustration of the effect of positioning a vehicle cargo area of interest on a vehicle's tail based on an embodiment of the present invention;
FIG. 8 is a flow chart illustrating a method for identifying objects of a vehicle having dangerous cargo-carrying behavior according to another embodiment of the present invention;
FIG. 9 is a schematic diagram of a vehicle object recognition device with dangerous cargo activity in accordance with one embodiment of the present invention;
FIG. 10 is a schematic illustration of a vehicle object recognition arrangement according to another embodiment of the present invention having dangerous cargo-carrying behavior;
FIG. 11 is a schematic diagram of a vehicle object recognition arrangement for dangerous cargo carrying behavior in accordance with yet another embodiment of the present invention;
fig. 12 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to accurately identify a vehicle target with dangerous cargo-carrying behaviors, the embodiment of the invention provides a vehicle target identification method and device with dangerous cargo-carrying behaviors and electronic equipment.
The terms in the examples of the present invention are explained as follows:
machine Learning: the method is a multi-field cross discipline and relates to a plurality of disciplines such as probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and the like. The special research on how a computer simulates or realizes the learning behavior of human beings so as to acquire new knowledge or skills and reorganize the existing knowledge structure to continuously improve the performance of the computer. It is the core of artificial intelligence and the fundamental way to make computer have intelligence.
Intelligent transportation: the comprehensive traffic transportation management system is established by effectively integrating and applying advanced information technology, data communication transmission technology, electronic sensing technology, control technology, computer technology and the like to the whole ground traffic management system, plays a role in a large range in all directions, and is real-time, accurate and efficient. The method is characterized in that information collection, processing, publishing, exchange, analysis and utilization are used as a main line, and diversified services are provided for traffic participants.
Dangerous cargo carrying behavior of the vehicle: the behavior that the vehicle carries cargo and has potential safety hazards is shown as follows: overloading the tricycle; truck overload cargo; car or bus roof suspension, lashing, etc. Such dangerous cargo activities are prone to traffic safety accidents.
CNN (Convolutional Neural Networks) is a machine learning model under deep supervised learning, and is the first learning algorithm that truly successfully trains a multi-layer network structure. It utilizes spatial relationships to reduce the number of parameters that need to be learned to improve the training performance of the general forward BP algorithm.
First, a method for identifying a vehicle target with dangerous cargo-carrying behavior according to an embodiment of the present invention will be described.
The execution main body of the method for recognizing the vehicle target with the dangerous cargo carrying behavior provided by the embodiment of the invention can be an electronic device executing an intelligent algorithm, the electronic device can be a camera (for example, an intelligent camera, a network camera and the like) executing the intelligent algorithm, and in order to realize the function of target recognition, the execution main body at least comprises a processor provided with a core processing chip. The method for identifying the vehicle target with dangerous cargo-carrying behavior provided by the embodiment of the invention can be realized in at least one mode of software, hardware circuit and logic circuit arranged in an execution body.
Fig. 2 is a schematic flow chart of a method for identifying a vehicle target with dangerous cargo-carrying behavior according to an embodiment of the present invention. The method mainly comprises the following steps: acquiring image data; acquiring vehicle detection information; positioning a vehicle cargo-carrying region of interest; and identifying whether dangerous cargo-carrying behaviors exist in the target vehicle.
The following describes in detail the steps of a method for identifying a vehicle object with dangerous cargo-carrying behavior according to an embodiment of the present invention, as shown in fig. 3.
S301, detecting the vehicle target of the image to be detected, and determining the area of the vehicle target in the image to be detected.
The image to be detected is an image needing to detect the vehicle target, whether the vehicle target with dangerous cargo-carrying behavior exists in the image or not is judged, and the position of the vehicle target with dangerous cargo-carrying behavior in the image is determined. The image to be detected can be a road traffic video image shot by a camera erected at a crossing, a detection station, a width-limiting (or height-limiting) bayonet and the like, and for the convenience of vehicle target detection, the image to be detected can be an image obtained by preprocessing the image shot by the camera through normalization, translation, overturning, segmentation and the like.
The method for detecting the vehicle target in the image to be detected can be a traditional method for comparing each target in the image to be detected with a vehicle target template, and determining the image to be detected as the vehicle target if the similarity is greater than a certain threshold; the method can also be used for detecting the neural network, the image to be detected is input into a neural network model which is trained in advance and used for detecting the vehicle target, and the output is the information of the vehicle target; the method can also be a multi-feature decision fusion method, wherein a front background separation is carried out on an image to be detected, a preliminary motion foreground block mass is extracted, then the motion foreground block mass is screened in the extracted motion foreground block mass to obtain a foreground region of interest, and then a final detection result is obtained based on the multi-feature decision fusion vehicle detection method. The above-mentioned manner of detecting the vehicle target in the image to be detected and other manners capable of detecting the vehicle target in the image all belong to the protection scope of the embodiment of the present invention, and are not limited specifically here. Through vehicle target detection, the obtained region where the vehicle target is located is often the minimum region containing all the features of the vehicle head or the vehicle tail, and the region may be a rectangular region, and may also be a circular region, a polygonal region, and the like.
S302, adjusting the area of the vehicle target according to the position relation of the area of the vehicle target in the image to be detected, and positioning the region of interest for carrying cargo of the vehicle in the image to be detected.
The region where the vehicle target is located is a region which only contains the head or the tail of the vehicle target in the image to be detected, and the region of interest for vehicle cargo loading comprises cargo-loading characteristics of the vehicle target. The vehicle with the dangerous cargo-carrying behavior is usually an overloaded vehicle, the peripheral outline of the vehicle with the dangerous cargo-carrying behavior is often larger than that of the vehicle with the safe cargo-carrying behavior, and of course, the vehicle with the dangerous cargo-carrying behavior may also be in the case that the loaded cargo is overlong, and the like.
The specific vehicle cargo area-of-interest positioning mode is shown in fig. 4, and mainly includes: detecting position coordinates of a vehicle target; calculating a position relation according to the position coordinates; and positioning the vehicle cargo area of interest according to the position relation.
Optionally, S302 may specifically be: determining a preset coordinate point in an area where a vehicle target is located; and amplifying the area where the vehicle target is located according to a preset proportion by taking the preset coordinate point as a reference to obtain the region of interest of the vehicle carrying cargo.
The preset coordinate point may be a center point of an area where the vehicle target is located, for example, if the area where the vehicle target is located is a circular area, the preset coordinate point may select a center of the circular area; if the area where the vehicle target is located is a rectangular area, the preset coordinate point can select the intersection point of the diagonal lines of the rectangular area; if the area where the vehicle target is located is a regular polygon area, the preset coordinate point can select the center of a circumscribed circle of the regular polygon area. The preset coordinate point may also be a certain point on a frame of an area where the vehicle target is located, for example, a vertex of the frame, a midpoint of the frame, or the like.
Amplifying the area where the vehicle target is located, namely actually expanding the area of the area where the vehicle target is located, for example, aiming at a circular area, taking the center of the circle as a reference, increasing the radius of the original circular area by a preset proportion, and obtaining a new concentric circle with a larger radius as a vehicle cargo-carrying region of interest; for another example, for a rectangular area, the length and width of the rectangular area may be increased by a preset ratio with respect to the center of the frame, and the rectangular area with a larger area may be obtained as the vehicle cargo area of interest.
Because the head and the tail of the vehicle target are generally rectangular and relatively few in other shapes, when the vehicle target is detected, the region where the obtained vehicle target is located is often a rectangular region. For the fact that the area where the vehicle target is located is a rectangular area, optionally, S302 may specifically be: the method comprises the steps of obtaining the width of a lower frame of an area where a vehicle target is located, wherein the area where the vehicle target is located is a rectangular area where the head or the tail of the vehicle target is located; the region of interest for vehicle cargo is located by setting a midpoint of the lower border as a lower border midpoint of the region of interest for vehicle cargo, setting a first height as a height of the region of interest for vehicle cargo, and setting a first width as a width of the region of interest for vehicle cargo, wherein the first height is a product of a width of the lower border and a first preset multiple, and the first width is a product of the width of the lower border and a second preset multiple.
Generally, the cargo carrying characteristics of the vehicle object are characterized in that the cargo carrying characteristics of the vehicle object are arranged on the top of the vehicle, the top of the vehicle object with dangerous cargo carrying behaviors is larger than that of a common vehicle object, so that the area can be enlarged to the top of the area where the vehicle object is located after the area where the vehicle object is located is detected through the vehicle object, and generally, in order to prevent the vehicle from being turned over, the cargo is arranged in a left-right symmetrical mode when the vehicle carries cargo, so that the area where the vehicle object is located can be enlarged based on the middle point of the lower frame of the area where the vehicle object is located.
Specifically, the distances from the middle point of the lower frame of the vehicle cargo area of interest to the upper frame and the left/right frames of the vehicle cargo area of interest can be calculated respectively by the formulas h, w, c, w, wherein h is the distance from the middle point of the lower frame of the vehicle cargo area of interest to the upper frame of the vehicle cargo area of interest, w is the distance from the middle point of the lower frame of the vehicle cargo area of interest to the left/right frames of the vehicle cargo area of interest, width is the width of the lower frame of the area where the vehicle target is located, b is a first preset multiple, c is a second preset multiple, and b and c can be values obtained by experimental statistics, for example, b is 1.2, and c is 0.6. And finally, calculating the vehicle cargo area of interest by combining the middle point of the lower frame of the vehicle cargo area of interest, the distance h from the middle point of the lower frame of the vehicle cargo area of interest to the upper frame of the vehicle cargo area of interest and the distance w from the middle point of the lower frame of the vehicle cargo area of interest to the left/right frames of the vehicle cargo area of interest. As shown in the dotted rectangular frame areas in fig. 5a and 5b, the solid rectangular frame area in the figure is the area where the vehicle target is located.
In locating a specific use of a vehicle cargo area of interest, as shown in fig. 6, it is first determined whether the vehicle detection result is valid, that is, it is necessary to determine whether the identified vehicle object is a valid vehicle object, and if not, it is not determined whether there is dangerous cargo behavior in the vehicle object.
Based on fig. 5a and 5b, the final positioning effect of the vehicle cargo region of interest is shown in fig. 7a and 7 b.
And S303, classifying and identifying the vehicle targets in the vehicle cargo carrying region of interest by adopting a pre-trained neural network model, and judging whether the vehicle targets are the vehicle targets with dangerous cargo carrying behaviors.
After the vehicle cargo-carrying interested region is obtained, a pre-trained neural network model can be adopted to identify the vehicle target with dangerous cargo-carrying behavior. Specifically, the vehicle targets in the vehicle cargo-carrying region of interest can be classified and identified through the CNN model, and then the result is output. Since it is generally only necessary to identify dangerous cargo-carrying behavior or no dangerous cargo-carrying behavior when identifying whether a vehicle object has dangerous cargo-carrying behavior, the CNN model may be selected as a binary model in general.
Optionally, the training process of the neural network model may include: obtaining a plurality of positive samples containing vehicle objects having dangerous cargo-carrying behavior and a plurality of negative samples containing vehicle objects not having dangerous cargo-carrying behavior; extracting each positive sample image and each negative sample image according to the coordinate information of the vehicle loading interested area containing the vehicle target in each calibrated positive sample and each calibrated negative sample, wherein the positive sample image is the image in the vehicle loading interested area in the positive sample, and the negative sample image is the image in the vehicle loading interested area in the negative sample; and training a neural network model based on the positive sample images and the negative sample images.
The trained neural network model is actually a classification model, can classify whether the vehicle target has the cargo-carrying behavior, can take a sample containing the vehicle target with the dangerous cargo-carrying behavior as a positive sample, and take a sample containing the vehicle target without the dangerous cargo-carrying behavior as a negative sample, and thus, the trained classification model can be a two-classification model. When positive samples and negative samples are obtained, theoretically, the more the samples are, the better the samples are, but the number of the vehicles which actually obtain dangerous cargo carrying behaviors is limited, so that a certain number can be met, but the number of the positive samples and the number of the negative samples are close to each other.
The image screenshot processing in the vehicle loading interested area is needed for the obtained positive and negative samples, the vehicle loading interested area can be determined based on the existing vehicle target area amplification mode, the amplification mode is not repeated, and the screenshot processing can be understood as extracting the image in the vehicle loading interested area through calibrating the coordinate information of the vehicle loading interested area, and training a neural network model based on the positive sample image and the negative sample image to obtain the binary model.
Whether a vehicle target with dangerous cargo-carrying behaviors exists in the image to be detected can be rapidly identified by inputting the image to be detected into a two-classification model, the position of the vehicle target is determined, the two-classification model is actually the operation of a plurality of convolution layers, the classification of the vehicle target is finally obtained through a classifier, the training process is the adjustment process of network layer parameters of each convolution layer, the output result of identifying the vehicle target with dangerous cargo-carrying behaviors is ensured, and the specific operation process is not repeated here.
And (4) inputting the graph 7a or the graph 7b into a pre-trained neural network model, so that a recognition result that the vehicle target has dangerous cargo-carrying behaviors can be obtained.
By applying the embodiment, the region where the vehicle target is located in the image to be detected is obtained by detecting the vehicle target in the image to be detected, the region where the vehicle target is located is adjusted according to the position relation of the region where the vehicle target is located in the image to be detected, the region of interest for vehicle cargo carrying containing the cargo carrying characteristics is located, and the vehicle target in the region of interest for vehicle cargo carrying is classified and identified by adopting the pre-trained neural network model, so that whether the vehicle target is the vehicle target with dangerous cargo carrying behavior or not can be judged. Compared with the target without the dangerous cargo-carrying behavior, the vehicle target with the dangerous cargo-carrying behavior has the advantages that the peripheral outline of the target is different in size, the region where the vehicle target is located is the region only containing the head or the tail of the vehicle target, so that the region where the vehicle target is located is adjusted to ensure that the adjusted region of interest for vehicle cargo-carrying can include the cargo-carrying characteristics of the vehicle target, and then the pre-trained neural network model is adopted, so that whether the vehicle target has the dangerous cargo-carrying behavior can be accurately classified and identified.
Based on the embodiment shown in fig. 3, the embodiment of the present invention further provides a method for identifying a vehicle object with dangerous cargo carrying behavior, and as shown in fig. 8, the method for identifying a vehicle object with dangerous cargo carrying behavior may include the following steps.
S801, detecting the vehicle target of the image to be detected, and determining the area of the vehicle target in the image to be detected.
S802, judging whether the size of the area where the vehicle target is located is larger than a preset size. If so, perform S803, otherwise perform S805.
And S803, adjusting the area where the vehicle target is located according to the position relation of the area where the vehicle target is located in the image to be detected, and positioning the region of interest for carrying cargo of the vehicle in the image to be detected.
S804, classifying and identifying the vehicle targets in the vehicle cargo-carrying interested region by adopting a pre-trained neural network model, and judging whether the vehicle targets are the vehicle targets with dangerous cargo-carrying behaviors.
And S805, discarding the area where the vehicle target is located.
When the camera is used for shooting a road image, the vehicle target is possibly far away from the camera, the area where the vehicle target is located is a very small area, and even if the area where the vehicle target is located is adjusted, whether the vehicle target is the vehicle target with dangerous cargo carrying behaviors or not cannot be accurately identified, therefore, a preset size is set, for example, 255 x 255, and when the size of the area where the vehicle target is located is larger than 255 x 255, the operation of area adjustment and identification of the vehicle target with dangerous cargo carrying behaviors can be carried out, so that the identification accuracy is guaranteed. If the size of the area where the vehicle target is located is smaller than or equal to the preset size, which indicates that the possibility of recognition error is high, the area where the vehicle target is located is discarded, namely, the operation of recognizing the vehicle target with dangerous cargo carrying behavior without area adjustment is carried out.
By applying the embodiment, the region where the vehicle target is located in the image to be detected is obtained by detecting the vehicle target in the image to be detected, the region where the vehicle target is located is adjusted according to the position relation of the region where the vehicle target is located in the image to be detected, the region of interest for vehicle cargo carrying containing the cargo carrying characteristics is located, and the vehicle target in the region of interest for vehicle cargo carrying is classified and identified by adopting the pre-trained neural network model, so that whether the vehicle target is the vehicle target with dangerous cargo carrying behavior or not can be judged. Compared with the target without the dangerous cargo-carrying behavior, the vehicle target with the dangerous cargo-carrying behavior has the advantages that the peripheral outline of the target is different in size, the region where the vehicle target is located is the region only containing the head or the tail of the vehicle target, so that the region where the vehicle target is located is adjusted to ensure that the adjusted region of interest for vehicle cargo-carrying can include the cargo-carrying characteristics of the vehicle target, and then the pre-trained neural network model is adopted, so that whether the vehicle target has the dangerous cargo-carrying behavior can be accurately classified and identified. And whether the operation of area adjustment and vehicle target identification with dangerous cargo carrying behaviors is carried out or not is determined by judging whether the size of the area where the detected vehicle target is located is larger than the preset size, so that the accuracy of target identification is ensured.
In accordance with the above method embodiment, the present invention provides a vehicle object recognition device with dangerous cargo-carrying behavior, and as shown in fig. 9, the vehicle object recognition device with dangerous cargo-carrying behavior may include the following modules.
The detecting module 910 is configured to perform vehicle target detection on an image to be detected, and determine an area where a vehicle target is located in the image to be detected.
The positioning module 920 is configured to adjust the area where the vehicle target is located according to a position relationship of the area where the vehicle target is located in the image to be detected, and position a vehicle cargo carrying region of interest in the image to be detected, where the area where the vehicle target is located is an area that only includes a head or a tail of the vehicle target in the image to be detected, and the vehicle cargo carrying region of interest includes cargo carrying features of the vehicle target.
The identifying module 930 is configured to perform classification and identification on the vehicle target in the vehicle cargo-carrying region of interest by using a pre-trained neural network model, and determine whether the vehicle target is a vehicle target with dangerous cargo-carrying behavior.
Optionally, the positioning module 920 may be specifically configured to: determining a preset coordinate point in an area where the vehicle target is located; and amplifying the area where the vehicle target is located according to a preset proportion by taking the preset coordinate point as a reference to obtain a vehicle cargo-carrying interested area containing cargo-carrying characteristics.
Optionally, the positioning module 920 may be specifically configured to: acquiring the width of a lower frame of an area where the vehicle target is located, wherein the area where the vehicle target is located is a rectangular area where the head or the tail of the vehicle target is located; the vehicle cargo area of interest is located by setting a midpoint of the lower frame to a lower frame midpoint of the vehicle cargo area of interest, setting a first height to a height of the vehicle cargo area of interest, and setting a first width to a width of the vehicle cargo area of interest, wherein the first height is a product of a width of the lower frame and a first preset multiple, and the first width is a product of a width of the lower frame and a second preset multiple.
By applying the embodiment, the vehicle target detection is carried out on the image to be detected to obtain the area where the vehicle target is located in the image to be detected, the area where the vehicle target is located is adjusted according to the position relation of the area where the vehicle target is located in the image to be detected, the vehicle cargo carrying interested area containing cargo carrying characteristics is located, the vehicle target in the vehicle cargo carrying interested area is classified and identified by adopting the pre-trained neural network model, and whether the vehicle target is the vehicle target with dangerous cargo carrying behaviors or not can be judged. Compared with a target without dangerous cargo-carrying behaviors, the vehicle target with dangerous cargo-carrying behaviors has different peripheral outline sizes, and the region where the vehicle target is located is the region only containing the head or the tail of the vehicle target, so that the region where the vehicle target is located is adjusted to ensure that the adjusted region of interest for vehicle cargo-carrying can include the cargo-carrying characteristics of the vehicle target, and then a pre-trained neural network model is adopted to accurately classify and identify whether the vehicle target has dangerous cargo-carrying behaviors.
Based on the embodiment shown in fig. 9, the embodiment of the invention also provides a vehicle object recognition device with dangerous cargo-carrying behavior, and as shown in fig. 10, the vehicle object recognition device with dangerous cargo-carrying behavior can comprise the following modules.
The detecting module 1010 is configured to perform vehicle target detection on an image to be detected, and determine an area where a vehicle target is located in the image to be detected.
The determining module 1020 is configured to determine whether the size of the area where the vehicle target is located is larger than a preset size.
A positioning module 1030, configured to, if the determination result of the determining module 1020 is yes, adjust the region where the vehicle target is located according to a position relationship of the region where the vehicle target is located in the image to be detected, and position a vehicle cargo-carrying region of interest in the image to be detected, where the region where the vehicle target is located is a region where only a head or a tail of the vehicle target is included in the image to be detected, and the vehicle cargo-carrying region of interest includes a cargo-carrying feature of the vehicle target.
The identifying module 1040 is configured to perform classification and identification on the vehicle targets in the vehicle cargo-carrying region of interest by using a pre-trained neural network model, and determine whether the vehicle targets are vehicle targets with dangerous cargo-carrying behaviors.
By applying the embodiment, the vehicle target detection is carried out on the image to be detected to obtain the area where the vehicle target is located in the image to be detected, the area where the vehicle target is located is adjusted according to the position relation of the area where the vehicle target is located in the image to be detected, the vehicle cargo carrying interested area containing cargo carrying characteristics is located, the vehicle target in the vehicle cargo carrying interested area is classified and identified by adopting the pre-trained neural network model, and whether the vehicle target is the vehicle target with dangerous cargo carrying behaviors or not can be judged. Compared with a target without dangerous cargo-carrying behaviors, the vehicle target with dangerous cargo-carrying behaviors has different peripheral outline sizes, and the region where the vehicle target is located is the region only containing the head or the tail of the vehicle target, so that the region where the vehicle target is located is adjusted to ensure that the adjusted region of interest for vehicle cargo-carrying can include the cargo-carrying characteristics of the vehicle target, and then a pre-trained neural network model is adopted to accurately classify and identify whether the vehicle target has dangerous cargo-carrying behaviors. And whether the operation of area adjustment and vehicle target identification with dangerous cargo carrying behaviors is carried out is determined by judging whether the size of the area where the detected vehicle target is located is larger than the preset size, so that the accuracy of target identification is ensured.
Based on the embodiment shown in fig. 9, the embodiment of the invention also provides a vehicle object recognition device with dangerous cargo carrying behavior, and as shown in fig. 11, the vehicle object recognition device with dangerous cargo carrying behavior may comprise the following modules.
The detecting module 1110 is configured to perform vehicle target detection on an image to be detected, and determine an area where a vehicle target is located in the image to be detected.
The positioning module 1120 is configured to adjust a region of the vehicle object according to a position relationship of the region of the vehicle object in the image to be detected, and position a region of interest for vehicle loading in the image to be detected, where the region of the vehicle object is a region that only includes a head or a tail of the vehicle object in the image to be detected, and the region of interest for vehicle loading includes a loading feature of the vehicle object.
The obtaining module 1130 obtains a plurality of positive samples containing vehicle objects having dangerous cargo activities and a plurality of negative samples containing vehicle objects not having dangerous cargo activities.
The extracting module 1140 is configured to extract the images of the positive samples and the images of the negative samples according to the calibrated coordinates of the region of interest for vehicle loading including the vehicle target in the positive samples and the negative samples, where the image of the positive sample is an image in the region of interest for vehicle loading in the positive sample, and the image of the negative sample is an image in the region of interest for vehicle loading in the negative sample.
A training module 1150, configured to train the neural network model based on the positive sample images and the negative sample images.
The identifying module 1160 is configured to perform classification and identification on the vehicle targets in the vehicle cargo-carrying region of interest by using a pre-trained neural network model, and determine whether the vehicle targets are vehicle targets with dangerous cargo-carrying behaviors.
By applying the embodiment, the vehicle target detection is carried out on the image to be detected to obtain the area where the vehicle target is located in the image to be detected, the area where the vehicle target is located is adjusted according to the position relation of the area where the vehicle target is located in the image to be detected, the vehicle cargo carrying interested area containing cargo carrying characteristics is located, the vehicle target in the vehicle cargo carrying interested area is classified and identified by adopting the pre-trained neural network model, and whether the vehicle target is the vehicle target with dangerous cargo carrying behaviors or not can be judged. Compared with the target without the dangerous cargo-carrying behavior, the vehicle target with the dangerous cargo-carrying behavior has the advantages that the peripheral outline of the target is different in size, the region where the vehicle target is located is the region only containing the head or the tail of the vehicle target, so that the region where the vehicle target is located is adjusted to ensure that the adjusted region of interest for vehicle cargo-carrying can include the cargo-carrying characteristics of the vehicle target, and then the pre-trained neural network model is adopted, so that whether the vehicle target has the dangerous cargo-carrying behavior can be accurately classified and identified.
An embodiment of the present invention further provides an electronic device, as shown in fig. 12, including a processor 1201 and a memory 1202, where,
a memory 1202 for storing computer programs;
the processor 1201 is configured to implement all the steps of the above-described method for identifying a vehicle target with dangerous cargo-carrying behavior when executing the computer program stored in the memory 1202.
The electronic device may be a Camera, and an image collector, such as an IPC (IP Camera), a smart Camera, or the like, may be further included in the electronic device.
The Memory may include a RAM (Random Access Memory) or an NVM (Non-Volatile Memory), such as at least one disk Memory. Optionally, the memory may also be at least one memory device located remotely from the processor.
The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also DSPs (Digital Signal Processing), ASICs (Application Specific Integrated circuits), FPGAs (Field-Programmable Gate arrays) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components.
In this embodiment, the processor of the camera can read the computer program stored in the memory and run the computer program to implement: the method comprises the steps of detecting a vehicle target in an image to be detected to obtain an area where the vehicle target is located in the image to be detected, adjusting the area where the vehicle target is located according to the position relation of the area where the vehicle target is located in the image to be detected, locating a vehicle cargo carrying interested area containing cargo carrying characteristics, and classifying and identifying the vehicle target in the vehicle cargo carrying interested area by adopting a pre-trained neural network model, so that whether the vehicle target is the vehicle target with dangerous cargo carrying behaviors or not can be judged. Compared with a target without dangerous cargo-carrying behaviors, the vehicle target with dangerous cargo-carrying behaviors has different peripheral outline sizes, and the region where the vehicle target is located is the region only containing the head or the tail of the vehicle target, so that the region where the vehicle target is located is adjusted to ensure that the adjusted region of interest for vehicle cargo-carrying can include the cargo-carrying characteristics of the vehicle target, and then a pre-trained neural network model is adopted to accurately classify and identify whether the vehicle target has dangerous cargo-carrying behaviors.
In addition, in response to the method for identifying an object of a vehicle with dangerous cargo-carrying behavior provided in the above embodiment, an embodiment of the present invention provides a machine-readable storage medium for storing a computer program, wherein when the computer program is executed by a processor, all steps of the method for identifying an object of a vehicle with dangerous cargo-carrying behavior are implemented.
In this embodiment, the machine-readable storage medium stores an application program that executes the method for identifying a vehicle object with dangerous cargo-carrying behavior provided by the embodiment of the present invention when running, so that the following can be implemented: the method comprises the steps of detecting a vehicle target in an image to be detected to obtain an area where the vehicle target is located in the image to be detected, adjusting the area where the vehicle target is located according to the position relation of the area where the vehicle target is located in the image to be detected, locating a vehicle cargo carrying interested area containing cargo carrying characteristics, and classifying and identifying the vehicle target in the vehicle cargo carrying interested area by adopting a pre-trained neural network model, so that whether the vehicle target is the vehicle target with dangerous cargo carrying behaviors or not can be judged. Compared with a target without dangerous cargo-carrying behaviors, the vehicle target with dangerous cargo-carrying behaviors has different peripheral outline sizes, and the region where the vehicle target is located is the region only containing the head or the tail of the vehicle target, so that the region where the vehicle target is located is adjusted to ensure that the adjusted region of interest for vehicle cargo-carrying can include the cargo-carrying characteristics of the vehicle target, and then a pre-trained neural network model is adopted to accurately classify and identify whether the vehicle target has dangerous cargo-carrying behaviors.
As for the embodiments of the electronic device and the machine-readable storage medium, since the contents of the related methods are substantially similar to those of the foregoing method embodiments, the description is relatively simple, and reference may be made to the partial description of the method embodiments for relevant points.
It should be noted that, in this document, relational terms such as first and second, and the like are 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 apparatus 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 apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on differences from other embodiments. In particular, for the apparatus, the electronic device, and the machine-readable storage medium embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and in relation to the description, reference may be made to some portions of the method embodiments.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (11)
1. A method of identifying objects in a vehicle having dangerous cargo-carrying behavior, the method comprising:
carrying out vehicle target detection on an image to be detected, and determining the area of a vehicle target in the image to be detected;
according to the position of the area where the vehicle target is located in the image to be detected, adjusting the area where the vehicle target is located, and positioning a vehicle loading interested area in the image to be detected, wherein the area where the vehicle target is located is an area which only contains the head or the tail of the vehicle target in the image to be detected, and the vehicle loading interested area comprises loading characteristics of the vehicle target;
And classifying and identifying the vehicle targets in the vehicle freight carrying interested region by adopting a pre-trained neural network model, and judging whether the vehicle targets are vehicle targets with dangerous freight carrying behaviors.
2. The method according to claim 1, wherein before the adjusting the region of the vehicle object according to the position of the region of the vehicle object in the image to be detected and locating the region of interest for vehicle cargo in the image to be detected, the method further comprises:
judging whether the size of the area where the vehicle target is located is larger than a preset size or not;
if so, adjusting the area of the vehicle target according to the position of the area of the vehicle target in the image to be detected, and positioning the region of interest for carrying cargo of the vehicle in the image to be detected.
3. The method according to claim 1 or 2, wherein the adjusting the area of the vehicle object according to the position of the area of the vehicle object in the image to be detected to locate the region of interest for vehicle cargo in the image to be detected comprises:
Determining a preset coordinate point in an area where the vehicle target is located;
and amplifying the area where the vehicle target is located according to a preset proportion by taking the preset coordinate point as a reference to obtain a vehicle cargo-carrying interested area.
4. The method according to claim 1 or 2, wherein the adjusting the area of the vehicle object according to the position of the area of the vehicle object in the image to be detected to locate the region of interest for vehicle cargo in the image to be detected comprises:
acquiring the width of a lower border of an area where the vehicle target is located, wherein the area where the vehicle target is located is a rectangular area where the head or the tail of the vehicle target is located;
the vehicle cargo area of interest is located by setting a midpoint of the lower frame to a lower frame midpoint of the vehicle cargo area of interest, setting a first height to a height of the vehicle cargo area of interest, and setting a first width to a width of the vehicle cargo area of interest, wherein the first height is a product of a width of the lower frame and a first preset multiple, and the first width is a product of a width of the lower frame and a second preset multiple.
5. The method of claim 1, wherein the training process of the neural network model comprises:
obtaining a plurality of positive samples containing vehicle objects having dangerous cargo-carrying behavior and a plurality of negative samples containing vehicle objects not having dangerous cargo-carrying behavior;
extracting each positive sample image and each negative sample image according to the coordinate information of the vehicle loading interested area containing the vehicle target in each calibrated positive sample and each calibrated negative sample, wherein the positive sample image is an image in the vehicle loading interested area in the positive sample, and the negative sample image is an image in the vehicle loading interested area in the negative sample;
and training the neural network model based on each positive sample image and each negative sample image.
6. A vehicle object recognition device for hazardous cargo carrying behavior, said device comprising:
the detection module is used for detecting a vehicle target in an image to be detected and determining the area of the vehicle target in the image to be detected;
the positioning module is used for adjusting the area where the vehicle target is located according to the position of the area where the vehicle target is located in the image to be detected, and positioning a vehicle loading interested area in the image to be detected, wherein the area where the vehicle target is located is an area which only contains the head or the tail of the vehicle target in the image to be detected, and the vehicle loading interested area comprises loading characteristics of the vehicle target;
And the recognition module is used for classifying and recognizing the vehicle targets in the vehicle cargo-carrying interested region by adopting a pre-trained neural network model and judging whether the vehicle targets are the vehicle targets with dangerous cargo-carrying behaviors or not.
7. The apparatus of claim 6, further comprising:
the judging module is used for judging whether the size of the area where the vehicle target is located is larger than a preset size or not;
and the positioning module is specifically used for adjusting the area where the vehicle target is located according to the position of the area where the vehicle target is located in the image to be detected if the judgment result of the judgment module is yes, and positioning the region of interest for vehicle cargo in the image to be detected.
8. The device according to claim 6 or 7, wherein the positioning module is specifically configured to:
determining a preset coordinate point in the area where the vehicle target is located;
and amplifying the region where the vehicle target is located according to a preset proportion by taking the preset coordinate point as a reference to obtain a vehicle cargo region of interest.
9. The device according to claim 6 or 7, wherein the positioning module is specifically configured to:
Acquiring the width of a lower frame of an area where the vehicle target is located, wherein the area where the vehicle target is located is a rectangular area where the head or the tail of the vehicle target is located;
the vehicle cargo area of interest is located by setting a midpoint of the lower frame to a lower frame midpoint of the vehicle cargo area of interest, setting a first height to a height of the vehicle cargo area of interest, and setting a first width to a width of the vehicle cargo area of interest, wherein the first height is a product of a width of the lower frame and a first preset multiple, and the first width is a product of a width of the lower frame and a second preset multiple.
10. The apparatus of claim 6, further comprising:
an acquisition module for acquiring a plurality of positive samples containing vehicle objects having dangerous cargo-carrying behavior and a plurality of negative samples containing vehicle objects not having dangerous cargo-carrying behavior;
the extraction module is used for extracting each positive sample image and each negative sample image according to the coordinate information of the vehicle loading interested area containing the vehicle target in each calibrated positive sample and each calibrated negative sample, wherein the positive sample image is an image in the vehicle loading interested area in the positive sample, and the negative sample image is an image in the vehicle loading interested area in the negative sample;
And the training module is used for training the neural network model based on each positive sample image and each negative sample image.
11. An electronic device comprising a processor and a memory, wherein,
the memory is used for storing a computer program;
the processor, when executing the computer program stored in the memory, is configured to perform the method steps of any of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810557515.XA CN110555347B (en) | 2018-06-01 | 2018-06-01 | Vehicle target identification method and device with dangerous cargo-carrying behavior and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810557515.XA CN110555347B (en) | 2018-06-01 | 2018-06-01 | Vehicle target identification method and device with dangerous cargo-carrying behavior and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110555347A CN110555347A (en) | 2019-12-10 |
| CN110555347B true CN110555347B (en) | 2022-06-03 |
Family
ID=68735275
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810557515.XA Active CN110555347B (en) | 2018-06-01 | 2018-06-01 | Vehicle target identification method and device with dangerous cargo-carrying behavior and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110555347B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111523417A (en) * | 2020-04-13 | 2020-08-11 | 北京巨视科技有限公司 | Truck overload identification method and device based on image matching |
| CN111507958B (en) * | 2020-04-15 | 2023-05-26 | 全球能源互联网研究院有限公司 | Target detection method, training method of detection model and electronic equipment |
| CN111985295A (en) * | 2020-06-05 | 2020-11-24 | 深圳市金溢科技股份有限公司 | Electric bicycle behavior recognition method and system, industrial personal computer and camera |
| CN111798698B (en) * | 2020-06-24 | 2022-04-08 | 中国第一汽车股份有限公司 | Method and device for determining front target vehicle and vehicle |
| CN112232449B (en) * | 2020-12-14 | 2021-04-27 | 浙江大华技术股份有限公司 | Neural network training method, electronic device, and storage medium |
| CN112598084B (en) * | 2021-03-02 | 2021-06-29 | 深圳金三立视频科技股份有限公司 | Vehicle type identification method and terminal based on image processing |
| JP7447870B2 (en) * | 2021-06-04 | 2024-03-12 | トヨタ自動車株式会社 | Information processing server, information processing server processing method, program |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101436253B (en) * | 2007-11-14 | 2012-04-25 | 东软集团股份有限公司 | Method and device for verifying interested area of vehicle |
| CN101996494A (en) * | 2009-08-31 | 2011-03-30 | 深圳市中交科科技有限公司 | Intelligent analysis and monitoring system of traffic images based on video discriminating technology |
| CN104537387A (en) * | 2014-12-16 | 2015-04-22 | 广州中国科学院先进技术研究所 | Method and system for classifying automobile types based on neural network |
| CN106295459A (en) * | 2015-05-11 | 2017-01-04 | 青岛若贝电子有限公司 | Based on machine vision and the vehicle detection of cascade classifier and method for early warning |
| CN105469052B (en) * | 2015-11-25 | 2018-10-19 | 东方网力科技股份有限公司 | A kind of vehicle detecting and tracking method and device |
| CN106114681B (en) * | 2016-08-30 | 2021-01-29 | 北京华力兴科技发展有限责任公司 | Vehicle parking early warning system and vehicle |
| CN106874863B (en) * | 2017-01-24 | 2020-02-07 | 南京大学 | Vehicle illegal parking and reverse running detection method based on deep convolutional neural network |
| CN206921292U (en) * | 2017-05-21 | 2018-01-23 | 曹进 | A kind of haulage vehicle violation supervisory systems |
| CN108061595A (en) * | 2017-12-14 | 2018-05-22 | 四川奇石缘科技股份有限公司 | A kind of traffic police administers the non-at-scene law enforcement detecting system of outline measuring overload and method |
-
2018
- 2018-06-01 CN CN201810557515.XA patent/CN110555347B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110555347A (en) | 2019-12-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110555347B (en) | Vehicle target identification method and device with dangerous cargo-carrying behavior and electronic equipment | |
| Chen et al. | Pothole detection using location-aware convolutional neural networks | |
| CN109816024B (en) | Real-time vehicle logo detection method based on multi-scale feature fusion and DCNN | |
| CN109087510B (en) | Traffic monitoring method and device | |
| CN107563372B (en) | License plate positioning method based on deep learning SSD frame | |
| CN109993031B (en) | Method and device for detecting target illegal driving behaviors of livestock-powered vehicle and camera | |
| Adu-Gyamfi et al. | Automated vehicle recognition with deep convolutional neural networks | |
| CN109190444B (en) | Method for realizing video-based toll lane vehicle feature recognition system | |
| CN110119726B (en) | Vehicle brand multi-angle identification method based on YOLOv3 model | |
| CN106295541A (en) | Vehicle type recognition method and system | |
| CN106250838A (en) | vehicle identification method and system | |
| CN114049572A (en) | Detection method for identifying small target | |
| Onim et al. | Traffic surveillance using vehicle license plate detection and recognition in bangladesh | |
| Achmad | Gaussian Mixture Models optimization for counting the numbers of vehicle by adjusting the Region of Interest under heavy traffic condition | |
| Arthi et al. | Object detection of autonomous vehicles under adverse weather conditions | |
| Chen | Road vehicle recognition algorithm in safety assistant driving based on artificial intelligence | |
| CN112766046A (en) | Target detection method and related device | |
| Zhang et al. | A front vehicle detection algorithm for intelligent vehicle based on improved gabor filter and SVM | |
| CN110634120B (en) | Vehicle damage judgment method and device | |
| CN113326831B (en) | Method and device for screening traffic violation data, electronic equipment and storage medium | |
| CN111126271B (en) | Bayonet snap image vehicle detection method, computer storage medium and electronic equipment | |
| Hasan Yusuf et al. | Real-Time Car Parking Detection with Deep Learning in Different Lighting Scenarios | |
| CN112232124A (en) | Crowd situation analysis method, video processing device and device with storage function | |
| Castillo et al. | Vsion: Vehicle occlusion handling for traffic monitoring | |
| CN114913488A (en) | Sprinkler detection method, device, electronic device, and storage medium |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |