CN113505767B - Cargo vehicle lane occupying operation behavior detection method and system based on cloud computing - Google Patents
Cargo vehicle lane occupying operation behavior detection method and system based on cloud computing Download PDFInfo
- Publication number
- CN113505767B CN113505767B CN202111058660.1A CN202111058660A CN113505767B CN 113505767 B CN113505767 B CN 113505767B CN 202111058660 A CN202111058660 A CN 202111058660A CN 113505767 B CN113505767 B CN 113505767B
- Authority
- CN
- China
- Prior art keywords
- customer
- behavior
- transaction
- road
- cargo vehicle
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 31
- 230000006399 behavior Effects 0.000 claims abstract description 123
- 238000000034 method Methods 0.000 claims abstract description 35
- 210000000707 wrist Anatomy 0.000 claims description 40
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000004590 computer program Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000005303 weighing Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a method and a system for detecting the road occupation operation behavior of a cargo vehicle based on cloud computing, and belongs to the technical field of road occupation operation behavior detection. The method comprises the following steps: acquiring continuous multi-frame images of a target area to be detected, and judging whether a cargo vehicle exists in the images or not; if yes, distinguishing the corresponding stall and customer of the cargo vehicle in the image; calculating the integrity of the set transaction logic sequence corresponding to each customer according to the behaviors of the stall owner and the customer in each frame of image; calculating the probability of the occurrence of the transaction behavior corresponding to each customer according to the integrity of the set transaction logic sequence corresponding to each customer; and judging whether the freight vehicle has the road occupation operation behavior or not according to the probability of the occurrence of the transaction behavior corresponding to each customer. The method judges whether the freight vehicle has the road-occupying operation behavior or not through the integrity of the set transaction logic sequence corresponding to each customer, and improves the accuracy of judging the road-occupying operation behavior of the freight vehicle.
Description
Technical Field
The invention relates to the technical field of lane occupation operation behavior detection, in particular to a method and a system for detecting the lane occupation operation behavior of a cargo vehicle based on cloud computing.
Background
At present, smart city projects are increasingly becoming the mainstream direction of city construction, and smart digital city management is an important component in smart city projects. The occupied road operation behavior is ubiquitous in city management affairs, and has multiple influences on the city development fields such as traffic safety, city appearance and city appearance. The influence of the road occupation operation behavior on traffic safety is particularly prominent, and the development of urban traffic is seriously hindered by traffic jam and even traffic accidents caused by the road occupation operation behavior.
The existing main way for detecting the lane occupation operation behavior is as follows: and obtaining suspicious static pictures, classifying objects in the pictures by using a deep learning network, and detecting the lane occupation operation behavior based on the object classification result. The above method has the following problems when detecting the occupied road operation:
1) the detection main body is the booth article and the road surface information, and the operation behavior of a mobile vendor taking a vehicle as a booth is not taken into consideration, so that the operation behavior of a cargo vehicle as a booth cannot be detected; 2) the method only detects a single-frame static image, does not consider time sequence information, and judges whether the road occupation operation behavior of the cargo vehicle exists only according to whether the vehicle carries goods and whether the vehicle stops at the roadside even if the operation behavior of the mobile vendor of the vehicle as a booth is considered, so that misjudgment is easy to occur.
Disclosure of Invention
The invention aims to provide a method and a system for detecting the lane occupation operation behavior of a cargo vehicle based on cloud computing, which are used for solving the problem that the existing detection method cannot accurately detect the lane occupation operation behavior of the cargo vehicle.
In order to solve the problems, the technical scheme of the method for detecting the road occupation operation behavior of the cargo vehicle based on the cloud computing comprises the following steps:
acquiring continuous multi-frame images of a target area to be detected, and judging whether a cargo vehicle exists in the images or not;
if yes, distinguishing the corresponding stall and customer of the cargo vehicle in the image;
calculating the integrity of the set transaction logic sequence corresponding to each customer according to the behaviors of the stall owner and the customer in each frame of image;
calculating the probability of the occurrence of the transaction behavior corresponding to each customer according to the integrity of the set transaction logic sequence corresponding to each customer;
and judging whether the freight vehicle has the road occupation operation behavior or not according to the probability of the occurrence of the transaction behavior corresponding to each customer.
The invention also provides a technical scheme of the system for detecting the road occupation operation behavior of the cargo vehicle based on the cloud computing, which comprises a memory and a processor, wherein the processor executes a computer program stored in the memory so as to realize the method for detecting the road occupation operation behavior of the cargo vehicle based on the cloud computing.
The detection method and the detection system have the beneficial effects that: the method obtains continuous multi-frame images of a target area to be detected, can obtain a series of booths and customer behaviors which are corresponding to the freight vehicles and have time sequence, judges the integrity of the set transaction logic sequence corresponding to the customers according to the behaviors of the booths and the customers corresponding to the freight vehicles, and indicates that the probability of the transaction behaviors of the customers and the booths is higher when the integrity of the set transaction logic sequence is higher; the invention takes whether the trading behavior of the stall owner and the customer is taken as the main reference factor for judging whether the road occupation of the cargo vehicle exists or not, judges whether the road occupation of the cargo vehicle exists or not by setting the integrity of the trading logic sequence corresponding to each customer, can avoid misjudgment of the trading behavior according to the behavior of the customer and the stall owner in a single frame image, namely, the misjudgment of the road occupation of the cargo vehicle, and improves the accuracy of judging the road occupation of the cargo vehicle.
Further, the method for determining the behavior of the booth and the customer in each frame of image comprises the following steps:
identifying target key points and target areas in the image, wherein the target key points comprise a stall owner wrist key point and a customer wrist key point, and the target areas comprise a payment area and a goods area;
and judging the behaviors of the stall owner and the customer according to the target key points and the target area.
Further, the method for determining the behavior of the stall owner and the customer according to the target key points and the target area comprises the following steps:
judging a first distance between a key point of a wrist of the customer in the image and a payment area, and judging whether the customer is paying by scanning a code according to the first distance;
judging a second distance between the key points of the wrists of the booth owner and the key points of the wrists of the customers in the image, and judging whether the customers exchange goods or pay cash with the booth owner according to the second distance;
and judging a third distance between the target key point in the image and the goods area, and judging whether the customer selects the goods or whether the stall helps the customer select the goods according to the third distance.
Further, the setting the transaction logic sequence includes:
customer picks goods-customer payment;
customer pick up goods-customer exchanges goods with the stall owner-customer pays;
the booth owner helps the customer to pick up the goods-the customer pays.
Further, the method for calculating the integrity of the set transaction logic sequence corresponding to each customer includes:
judging whether the behaviors of the booth owner and the customer in each frame of image belong to the set transaction sub-behaviors or not;
if so, arranging the behaviors of the stall owner and the customer according to the time sequence to obtain a transaction sub-behavior time sequence logic sequence;
and calculating the integrity of the set transaction logic sequence corresponding to the customer according to the transaction sub-behavior time sequence logic sequence and the matched set transaction logic sequence.
Further, the method for judging whether the freight vehicle has the road occupation operation behavior comprises the following steps:
calculating the probability of the corresponding transaction behavior of the cargo vehicle according to the probability of the corresponding transaction behavior of each customer;
and judging whether the probability of the corresponding transaction behavior of the cargo vehicle is greater than a set probability threshold, and if so, judging that the cargo vehicle has the lane occupying behavior.
Further, the probability of the existence of the transaction behavior corresponding to each customer is calculated by adopting the following calculation formula:
wherein,
the probability of the existence of a transaction for the ith customer,
the average value of the maximum values of the coincidence state probabilities corresponding to the transaction sub-behaviors corresponding to the ith customer is shown,
represents the ith customer pairThe integrity of the transaction logic sequence should be set.
Further, the owner and the customer corresponding to the cargo vehicle in the image are distinguished according to the frequency of occurrence of each person in the image.
Drawings
FIG. 1 is a flow chart of a method for detecting the lane-occupied operation of a cargo vehicle based on cloud computing according to the invention;
FIG. 2 is a schematic diagram of a set transaction logic sequence of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described below with reference to the drawings in the embodiments of the present invention.
Embodiment of method for detecting road occupation operation behavior of cargo vehicle based on cloud computing
The embodiment aims to realize accurate detection of the occupation behavior of the cargo vehicle, and as shown in fig. 1, the method for detecting the occupation operation behavior of the cargo vehicle based on cloud computing comprises the following steps:
1) acquiring continuous multi-frame images of a target area to be detected, and judging whether a cargo vehicle exists in the images or not;
in the embodiment, the video image of the target area to be detected is acquired by using the camera, and when the acquired video image includes a vehicle, whether the vehicle is a cargo-carrying vehicle can be judged by judging whether the vehicle includes a cargo-carrying part, such as a cargo hopper. The embodiment collects images through the camera and processes the collected images to realize the detection of the lane occupation management of the cargo vehicle, and due to the limited computing capability of the general camera, the embodiment uploads the collected images to the cloud server side for subsequent related computation after the cargo vehicle is detected, so that the reasonable implementation of the scheme is ensured.
The cameras in the embodiment are multiple cameras which are laid in a region where lane occupation is frequently managed in a smart city project, and the spatial positions of all points in a target region to be detected can be obtained by calibrating and pose calibrating the double cameras and reconstructing a three-dimensional space through binocular stereo vision. Since a video image is also a collection of continuous multi-frame images per se, as another embodiment, a camera may be used to acquire continuous multi-frame images instead of a video image.
The embodiment performs semantic segmentation on the acquired image based on the DNN network, can extract vehicle information, cargo information, pedestrian information and background information in the image, and can realize the judgment of whether the acquired image comprises vehicles or not. The DNN network for implementing semantic segmentation is built in an edge device, i.e., a camera, and performing semantic segmentation on an image based on the DNN network is the prior art, and details of performing semantic segmentation on an image based on the DNN network are not described herein again.
When the cargo carrying vehicle is not judged to exist in the image, the judgment of whether the cargo carrying vehicle exists in the image is only needed to be carried out continuously; when the cargo vehicle is judged to be present in the image, whether the cargo vehicle has the operation behavior of occupying the road is judged through the following steps.
2) If yes, distinguishing the corresponding stall and customer of the cargo vehicle in the image;
in the embodiment, the camera starts to acquire continuous multi-frame images of the target area to be detected from the starting time point, uploads the images to the cloud computing platform after the goods-carrying vehicles exist in the images, obtains the goods area S from the edges of the goods identified by the DNN network at the cloud server side, and expands the goods area S into the goods area S
Will be
And recording as a transaction behavior detection area. In this embodiment, the edge of the cargo area S is specifically extended by 2 meters to obtain the transaction behavior detection area
As other embodiments, the transaction behavior detection area can be determined according to the requirement
。
Detecting regions based on transaction behavior
The occurrence frequency of each person in the cloud server is used for distinguishing a spreader and a customer, specifically, a multi-frame image is obtained from the starting time point, the persons appearing in the transaction behavior detection area are matched frame by frame at the cloud server, the occurrence frequency of the same person in the multi-frame image is recorded, the label of the person with the highest occurrence frequency is marked as the spreader, and other persons are marked as customers.
The embodiment detects the area through the transaction behavior
The occurrence frequency of Chinese people can distinguish the stall and the customer, and as other implementation modes, the detection area of the transaction behavior of each person can be used
The time of occurrence of the above-mentioned steps is used for distinguishing the stall owner from the customer, and the stall owner is generally in the transaction behavior detection area
The time length of the occurrence of the customer is longer than the transaction behavior detection area of each customer
The length of time of occurrence of (c).
3) Calculating the integrity of the set transaction logic sequence corresponding to each customer according to the behaviors of the stall owner and the customer in each frame of image;
if the cargo vehicle is in a lane operation, the booth owner and the customer may have the following actions: the customer scans the code/pays the cash, the customer gives the selected goods to the spreader for weighing, the spreader returns the weighed goods to the customer, the spreader helps the customer select the goods, the customer selects the goods, and the like. In the embodiment, the integrity of the set transaction logic sequence corresponding to each customer is determined by detecting whether the above behaviors exist in the booth and the customer in the image and the sequence of the above behaviors, so as to determine whether the transaction behavior occurs to each customer by the integrity of the set transaction logic sequence corresponding to each customer.
In order to realize the judgment of whether the booths and the customers have the behaviors, the cloud server side marks the wrist key points of the booths and the customers in the image and identifies the payment area and the goods area in the image, and the behavior of the booths and the customers is judged through the marked wrist key points of the booths and the customers, the identified payment area and the identified goods area, wherein the method for identifying the payment area at the cloud server side of the embodiment is as follows: the method comprises the steps of calibrating the wrist key point position of each customer in a transaction behavior detection area frame by frame from a starting time point, calibrating customer tags belonging to the positions on a background, generating a customer tag quantity heat map by taking the number of the customer tags as heat, wherein the wrist key point of each customer is only possibly frequent in a goods area and a payment area, so that the area with the maximum heat except the goods area in the transaction behavior detection area is the payment area, such as an area where a payment two-dimensional code is pasted or placed, and the central point of the corresponding payment area is marked as H in the embodiment. In another embodiment, the area in the image where the payment two-dimensional code is pasted or placed may be identified, and the identified area may be used as the payment area.
In the embodiment, the method for judging the behaviors of the stall owner and the customer on the basis of the wrist key points of the stall owner and the customer, the identified payment area and the identified goods area on the cloud server side comprises the following steps:
determining a first distance d1 between the key point of the customer's wrist and the payment area in the image, and determining whether the customer is paying in the code scanning mode according to the first distance d 1; in this embodiment, the first distance d1 is calculated by finding the distance between the calibrated key point of the customer's wrist and the center point H of the code scanning area, and when the first distance d1 is smaller than a first set threshold, it is determined that the key point of the customer's wrist and the code scanning area coincide, and it is determined that the customer is scanning the code;
determining a second distance d2 between the key points of the stall owner's wrist and the key points of the customer's wrist in the image, and determining whether the customer is exchanging goods or paying cash with the stall owner according to the second distance d 2; in this embodiment, when the second distance is smaller than the second set threshold, it is determined that the key point of the wrist of the booth owner coincides with the key point of the wrist of the customer, and it is determined that the customer exchanges goods with the booth owner or pays cash;
judging a third distance d3 between the key point of the wrist of the customer or the key point of the wrist of the stall owner and the goods area in the image, and judging whether the customer selects the goods or whether the stall owner helps the customer select the goods according to the third distance d 3; in this embodiment, the shortest distance between the key point of the wrist of the customer or the key point of the wrist of the stall owner and the edge of the goods area is specifically taken as the third distance d3, and when the third distance d3 is smaller than the third set threshold, it is determined that the key point of the wrist of the customer or the key point of the wrist of the stall owner coincides with the code scanning goods area, and it is determined that the customer is picking up goods or the stall owner is helping the customer to pick up goods.
The behavior of selecting goods by a customer is defined as a transaction sub-behavior 1, the behavior of submitting the goods to a spreader for weighing by the customer is defined as a transaction sub-behavior 2, the behavior of receiving the goods submitted back by the spreader by the customer is defined as a transaction sub-behavior 3, the behavior of paying by the customer in a code scanning mode is defined as a transaction sub-behavior 4, the behavior of selecting the goods by the spreader side customer is defined as a transaction sub-behavior 5, and the behavior of paying by cash by the customer is defined as a transaction sub-behavior 6. As shown in fig. 2, the transaction action sequence generally includes the following: 1-2-3-4 (3, 4 may be in order), 1-2-3-6 (3, 6 may be in order), 1-4, 1-6, 1-5-3-4 (3, 4 may be in order), 1-5-3-6 (3, 6 may be in order), 5-3-4 (3, 4 may be in order), 5-3-6 (3, 6 may be in order).
The method for calculating the integrity of the set transaction logic sequence corresponding to each customer at the cloud server side in the embodiment comprises the following steps:
setting the time when a certain customer enters the transaction behavior detection area as the system starting time as T0;
arranging all detected sub-behaviors of the customer and the stall owner in a detection time sequence to obtain a transaction sub-behavior time sequence logic sequence, for example: the obtained corresponding arrangement of the customers is 1-3-4; and detecting the child behaviors of the same customer in a time sequence, and when the time sequence is found to be inconsistent with the standard logic time sequence, eliminating the behaviors which are not consistent with the time sequence, and only keeping the behaviors with correct time sequences.
And comparing the obtained transaction sub-behavior time sequence logic sequence with the set transaction logic sequence by taking the time Te when the customer leaves the detection area as the end time, and setting the one with the largest intersection with the obtained transaction sub-behavior time sequence logic sequence in the transaction logic sequence as a standard transaction logic sequence by using standard logic.
Defining the integrity of the set transaction logic sequence corresponding to each customer as
The calculation method is as follows:
wherein,
the number is given to the customer for the number,
a set of transaction sub-behaviors for which the correspondingly numbered customer is detected to be present in the transaction behavior detection area,
for the corresponding customer
The set of transaction sub-behaviors of (a) corresponds to the set of transaction sub-behaviors of the standard transaction logic order,
to represent
And
the number of categories of sub-behaviors of the same transaction,
to represent
The number of transaction sub-behavior categories. For example: the detected child behavior corresponding to the customer is arranged as 1-3-4, and then the corresponding set transaction logic sequence is 1-2-3-4 or 1-5-3-4,
，
the integrity of the set transaction logic sequence corresponding to the customer is 0.75.
4) Calculating the probability of the occurrence of the transaction behavior corresponding to each customer according to the integrity of the set transaction logic sequence corresponding to each customer;
in order to avoid the situation that the coincidence state cannot be accurately judged due to inaccurate threshold setting, for example, when the threshold setting is large, two distances may be smaller than the corresponding thresholds, the coincidence state probability value of the customer wrist key point and the booth owner wrist key point, the payment area or the goods area is calculated by the following formula, so as to represent the probability of a certain transaction child behavior of the booth owner and the customer by the coincidence state probability value:
wherein,
representing the customer's wrist key points and
the probability value of the coincidence state of (2),
to calculate
The intermediate variable utilized, j =1,2, 3; when j =1, the signal is transmitted,
representing the probability value of the coincidence state of the key point of the wrist of the customer and the code scanning area; when j =2, the signal is transmitted,
representing the probability value of the coincidence state of the customer wrist key point and the stall owner wrist key point; when j =3, the signal is transmitted,
and representing the probability value of the coincidence state of the key point of the wrist of the customer and the goods area.
Similarly, the distance between the key point of the stall owner's wrist and the key point of the customer's wrist, the distance between the key point of the stall owner's wrist and the code scanning area, and the distance between the key point of the stall owner's wrist and the goods area can be calculated according to the formula, and then the probability value of the coincidence state of the key point of the stall owner's wrist to the goods area can be calculated according to the formula
。
Each frame image corresponds to a group of the coincidence probability values, however, considering that the occurrence of a certain transaction sub-behavior is represented as long as the transaction sub-behavior is detected in a single frame image, the embodiment detects the change situation of the coincidence state probability values according to the corresponding time sequence of the acquired multi-frame image, and judges that the coincidence state probability value increases along with the timeCutting off the corresponding coincidence state probability value of each frame in the next period of time, taking the maximum value of the coincidence probability value as the probability of the corresponding transaction sub-behavior in the period of time, and judging that the transaction sub-behavior exists when the probability of the transaction sub-behavior is greater than a set probability threshold, wherein the probability of the corresponding transaction sub-behavior is to be determined
The corresponding maximum value is recorded as
Will be
The corresponding maximum value is recorded as
Will be
The corresponding maximum value is recorded as
Will be
The corresponding maximum value is recorded as
。
On the basis of determining the transaction sub-behaviors existing in the image and obtaining the occurrence probability of each transaction sub-behavior, the present embodiment uses the average value of the occurrence probabilities of the corresponding transaction sub-behaviors at the cloud server end to calculate the probability of the transaction behavior existing in the customer, specifically:
wherein,
the detected average value of the maximum value of the probability of the coincidence state corresponding to each transaction sub-behavior of the customer i,
indicating the probability that customer i has a transaction activity,
for customers
And correspondingly setting the integrity of the transaction logic sequence. For example: the detected corresponding child behaviors of the customer are arranged into 1-3-4, and then
Is just as
、
And
is measured.
The embodiment also finds undetected transaction sub-behaviors based on the compared standard transaction logic sequence, that is, performs temporal re-detection on the video data. If the missing sub-behaviors can be detected in the corresponding time sequence direction, determining the maximum value of the coincidence state probability corresponding to the transaction sub-behaviors and recalculating the existence probability of the transaction behaviors; if not, directly outputting the probability
。
5) And judging whether the freight vehicle has the road occupation operation behavior or not according to the probability of the occurrence of the transaction behavior corresponding to each customer.
According to the probability of the corresponding transaction behavior of each customer, calculating the probability of the corresponding transaction behavior of the cargo carrying vehicle at the cloud server side, wherein the calculation formula is as follows:
wherein m is the detected total number of the customers, and W is the probability of the corresponding transaction behavior of the cargo vehicle.
After the probability of the existence of the transaction behavior corresponding to the cargo vehicle is obtained, whether the probability of the existence of the transaction behavior corresponding to the cargo vehicle is larger than a set probability threshold value or not is further judged, and if the probability is larger than the set probability threshold value, the existence of the road occupation behavior of the cargo vehicle is judged.
As another embodiment, the probability of the existence transaction behavior corresponding to the cargo vehicle may not be calculated, and the existence of the vehicle occupying the road operation behavior of the cargo vehicle may be determined as long as the probability of the occurrence of the transaction behavior of more than a set number of customers is greater than a set probability.
In the embodiment, whether the trading behavior of the stall owner and the customer occurs is used as a main reference factor for judging whether the road occupation of the cargo vehicle exists, whether the road occupation of the cargo vehicle exists is judged according to the integrity of the set trading logic sequence corresponding to each customer, the misjudgment of the trading behavior of the customer and the stall owner in a single-frame image can be avoided, the misjudgment of the road occupation of the cargo vehicle is also avoided, and the accuracy of the judgment of the road occupation of the cargo vehicle is improved.
Embodiment of cargo vehicle lane occupying operation behavior detection system based on cloud computing
The cloud-computing-based road-occupation operation behavior detection system for the cargo vehicle comprises a memory and a processor, wherein the processor executes a computer program stored in the memory to realize the cloud-computing-based road-occupation operation behavior detection method for the cargo vehicle as described in the cloud-computing-based road-occupation operation behavior detection method embodiment.
In the embodiment of the method for detecting the road occupation operation of the cargo vehicle based on the cloud computing, the method for detecting the road occupation operation of the cargo vehicle based on the cloud computing has been described, and thus the description is omitted here.
It should be noted that while the preferred embodiments of the present invention have been described, additional variations and modifications to these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
Claims (9)
1. A cargo vehicle road occupation operation behavior detection method based on cloud computing is characterized by comprising the following steps:
acquiring continuous multi-frame images of a target area to be detected, and judging whether a cargo vehicle exists in the images or not;
if yes, distinguishing the corresponding stall and customer of the cargo vehicle in the image;
calculating the integrity of the set transaction logic sequence corresponding to each customer according to the behaviors of the stall owner and the customer in each frame of image;
calculating the probability of the occurrence of the transaction behavior corresponding to each customer according to the integrity of the set transaction logic sequence corresponding to each customer;
and judging whether the freight vehicle has the road occupation operation behavior or not according to the probability of the occurrence of the transaction behavior corresponding to each customer.
2. The cloud-computing-based method for detecting the road-occupation management behavior of a cargo vehicle according to claim 1, wherein the method for determining the behavior of the booth and the customer in each image frame comprises:
identifying target key points and target areas in the image, wherein the target key points comprise a stall owner wrist key point and a customer wrist key point, and the target areas comprise a payment area and a goods area;
and judging the behaviors of the stall owner and the customer according to the target key points and the target area.
3. The cloud-computing-based method for detecting the road-occupation operation of a cargo vehicle according to claim 2, wherein the method for determining the behavior of the booth owner and the customer according to the target key points and the target area comprises:
judging a first distance between a key point of a wrist of the customer in the image and a payment area, and judging whether the customer is paying by scanning a code according to the first distance;
judging a second distance between the key points of the wrists of the booth owner and the key points of the wrists of the customers in the image, and judging whether the customers exchange goods or pay cash with the booth owner according to the second distance;
and judging a third distance between the target key point in the image and the goods area, and judging whether the customer selects the goods or whether the stall helps the customer select the goods according to the third distance.
4. The cloud-computing-based method for detecting the road-occupation operation of a utility vehicle according to claim 1, wherein the setting of the transaction logic sequence comprises:
customer picks goods-customer payment;
customer pick up goods-customer exchanges goods with the stall owner-customer pays;
the booth owner helps the customer to pick up the goods-the customer pays.
5. The cloud-computing-based method for detecting the road-occupying operation behavior of the cargo vehicle according to claim 1, wherein the method for calculating the integrity of the set transaction logic sequence corresponding to each customer comprises the following steps:
judging whether the behaviors of the booth owner and the customer in each frame of image belong to the set transaction sub-behaviors or not;
if so, arranging the behaviors of the stall owner and the customer according to the time sequence to obtain a transaction sub-behavior time sequence logic sequence;
and calculating the integrity of the set transaction logic sequence corresponding to the customer according to the transaction sub-behavior time sequence logic sequence and the matched set transaction logic sequence.
6. The cloud-computing-based method for detecting the road-occupation operation behavior of the commercial vehicle according to claim 1, wherein the method for determining whether the road-occupation operation behavior exists comprises the following steps:
calculating the probability of the corresponding transaction behavior of the cargo vehicle according to the probability of the corresponding transaction behavior of each customer;
and judging whether the probability of the corresponding transaction behavior of the cargo vehicle is greater than a set probability threshold, and if so, judging that the cargo vehicle has the lane occupying behavior.
7. The cloud-computing-based method for detecting the road-occupying operation behavior of the cargo vehicle according to claim 1, wherein the probability of the existence of the transaction behavior corresponding to each customer is calculated by using the following calculation formula:
wherein,
the probability of the existence of a transaction for the ith customer,
the mean value of the maximum value of the probability of the coincidence state corresponding to each transaction sub-behavior corresponding to the ith customer is shown,
and the integrity of the set transaction logic sequence corresponding to the ith customer is shown.
8. The cloud-computing-based method for detecting the road-occupied business behavior of the cargo vehicle according to claim 1, wherein the owner and the customer of the cargo vehicle in the image are distinguished according to the frequency of occurrence of each person in the image.
9. A cloud-computing-based road-occupation management behavior detection system for a utility vehicle, comprising a memory and a processor, wherein the processor executes a computer program stored in the memory to implement the cloud-computing-based road-occupation management behavior detection method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111058660.1A CN113505767B (en) | 2021-09-10 | 2021-09-10 | Cargo vehicle lane occupying operation behavior detection method and system based on cloud computing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111058660.1A CN113505767B (en) | 2021-09-10 | 2021-09-10 | Cargo vehicle lane occupying operation behavior detection method and system based on cloud computing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113505767A CN113505767A (en) | 2021-10-15 |
| CN113505767B true CN113505767B (en) | 2021-11-16 |
Family
ID=78017073
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111058660.1A Active CN113505767B (en) | 2021-09-10 | 2021-09-10 | Cargo vehicle lane occupying operation behavior detection method and system based on cloud computing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113505767B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116597295A (en) * | 2023-07-19 | 2023-08-15 | 北京大也智慧数据科技服务有限公司 | Method and device for identifying lane occupation operation behavior |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2889842A1 (en) * | 2013-12-31 | 2015-07-01 | Patents Factory Ltd. Sp. z o.o. | A method for estimating dynamics of motion in a video image |
| CN109345435A (en) * | 2018-12-07 | 2019-02-15 | 山东晴天环保科技有限公司 | Occupy-street-exploit managing device and method |
| CN109978668A (en) * | 2019-04-02 | 2019-07-05 | 苏州工业职业技术学院 | Based on the matched transaction customer recognition methods of timestamp, system, equipment and medium |
| CN110334965A (en) * | 2019-07-12 | 2019-10-15 | 重庆工商职业学院 | A kind of huckster management method based on big data |
| CN113095301A (en) * | 2021-05-21 | 2021-07-09 | 南京甄视智能科技有限公司 | Road occupation operation monitoring method, system and server |
-
2021
- 2021-09-10 CN CN202111058660.1A patent/CN113505767B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2889842A1 (en) * | 2013-12-31 | 2015-07-01 | Patents Factory Ltd. Sp. z o.o. | A method for estimating dynamics of motion in a video image |
| CN109345435A (en) * | 2018-12-07 | 2019-02-15 | 山东晴天环保科技有限公司 | Occupy-street-exploit managing device and method |
| CN109978668A (en) * | 2019-04-02 | 2019-07-05 | 苏州工业职业技术学院 | Based on the matched transaction customer recognition methods of timestamp, system, equipment and medium |
| CN110334965A (en) * | 2019-07-12 | 2019-10-15 | 重庆工商职业学院 | A kind of huckster management method based on big data |
| CN113095301A (en) * | 2021-05-21 | 2021-07-09 | 南京甄视智能科技有限公司 | Road occupation operation monitoring method, system and server |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113505767A (en) | 2021-10-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9129524B2 (en) | Method of determining parking lot occupancy from digital camera images | |
| CN109087510B (en) | Traffic monitoring method and device | |
| CN102803991B (en) | Object detection device | |
| Schindler et al. | Automatic detection and tracking of pedestrians from a moving stereo rig | |
| US20130226667A1 (en) | Methods and apparatus to analyze markets based on aerial images | |
| US10102431B2 (en) | Visual monitoring of queues using auxillary devices | |
| KR101960899B1 (en) | Method for recognizing products | |
| WO2014151681A2 (en) | Methods and apparatus to estimate demography based on aerial images | |
| US10586115B2 (en) | Information processing device, information processing method, and computer program product | |
| CN112668525B (en) | People flow counting method and device, electronic equipment and storage medium | |
| CN111783589B (en) | Complex scene crowd counting method based on scene classification and multi-scale feature fusion | |
| US20140028447A1 (en) | System and Method of Tracking Vehicles Within a Parking Lot Using RFID Tags | |
| CN111898491A (en) | Method and device for identifying reverse driving of vehicle and electronic equipment | |
| Liu et al. | Vehicle detection and ranging using two different focal length cameras | |
| Tak et al. | Development of AI‐Based Vehicle Detection and Tracking System for C‐ITS Application | |
| CN113505767B (en) | Cargo vehicle lane occupying operation behavior detection method and system based on cloud computing | |
| CN115862043A (en) | System and method for merchandise identification | |
| CN112800841B (en) | Pedestrian counting method, device and system and computer readable storage medium | |
| CN117876994A (en) | Parking space recognition model training method, parking space recognition method and related device | |
| CN111598158B (en) | Warehouse inventory state sensing method based on AI analysis technology | |
| CN112365315A (en) | Commodity display position recommendation method, device and system and storage medium | |
| Chen et al. | Vision-based traffic surveys in urban environments | |
| CN104463135A (en) | Vehicle logo recognition method and system | |
| CN115131726A (en) | Parking space detection method, device, equipment and medium | |
| JP2019174910A (en) | Information acquisition device and information aggregation system and information aggregation device |
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 |