CN106682590B - Processing method of monitoring service and server - Google Patents
Processing method of monitoring service and server Download PDFInfo
- Publication number
- CN106682590B CN106682590B CN201611117138.5A CN201611117138A CN106682590B CN 106682590 B CN106682590 B CN 106682590B CN 201611117138 A CN201611117138 A CN 201611117138A CN 106682590 B CN106682590 B CN 106682590B
- Authority
- CN
- China
- Prior art keywords
- processing result
- service
- end equipment
- server
- scene
- 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
Classifications
-
- 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
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/94—Hardware or software architectures specially adapted for image or video understanding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/98—Detection or correction of errors, e.g. by rescanning the pattern or by human intervention; Evaluation of the quality of the acquired patterns
- G06V10/993—Evaluation of the quality of the acquired pattern
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- 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
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Software Systems (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
- Image Analysis (AREA)
Abstract
The application discloses a processing method of monitoring service, a server receives a first processing result of a designated service, which is reported by front-end equipment, and judges whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value or not through a preset confidence coefficient evaluation model; if the judgment result is yes, the server takes the first processing result as the processing result of the appointed service; if the judgment result is negative, the server processes the appointed service, a second processing result is obtained, and the second processing result is used as the processing result of the appointed service. The first processing result of the front-end equipment is evaluated through the server, when the confidence coefficient of the first processing result is higher, the first processing result is adopted, and when the confidence coefficient of the first processing result is lower, the server further processes the appointed service, so that the resources of the front-end equipment are fully utilized on the basis of ensuring the service processing accuracy, and the comprehensive resource utilization rate of the monitoring system is improved.
Description
Technical Field
The application relates to the technical field of communication, in particular to a processing method of monitoring service, and also particularly relates to a server.
Background
At present, video monitoring is advanced to high definition and intelligent. Intelligent monitoring has been applied in numerous fields, such as intelligent traffic, intelligent parks, safe cities, etc. With the development of computer image recognition, video intelligent algorithms, such as technologies of motion detection, forbidden zone intrusion, tripwire warning, people counting, license plate recognition, pedestrian recognition, face detection, recognition and the like, are increasingly demanded. The intelligent algorithms can be widely applied to various scenes, such as a park entrance, a key area, a square, a subway, a sidewalk, a city, a highway and the like.
Common city streets, overhead, parks, building entrances and exits, outdoor scene target detection scenes, people and vehicles mixed, daytime and night, shadow changes such as rainy days, ambient light reflection, and other scene light changes, and all the interference factors have adverse effects on target detection. To address these disturbances, more intelligent analysis algorithms to reject the disturbance are required, which means more system computing resources and other system resources are required.
Most of the existing intelligent analysis algorithm schemes are server schemes, and target detection, tracking, analysis and recognition are carried out on the accessed monitoring video and picture, so that a target recognition analysis result is output. And generating an alarm by combining the defense deployment rule or retrieving the target according to the retrieval rule.
In addition, with the improvement of the performance of the front-end equipment chip, intelligent services have been partially migrated to the front-end equipment. But the performance and resources of the front-end chip are still relatively lower than those of the back-end server, the functions are not rich, the accuracy index is not high, and more false positives and false negatives exist.
In the process of realizing the application, the inventor finds that the existing intelligent monitoring service has at least the following problems:
1. the algorithm which can be deployed by simply relying on the computing resources of the front-end equipment is limited, and the requirement of accurate detection is difficult to meet.
2. The method simply relies on the server at the back end to process, so that a large number of servers are required to be deployed at the back end, the interference problem is solved, the target fine characteristic identification information is provided, and the accuracy of target detection and identification is improved. However, the price of the back-end server is high, so that the cost performance of the scheme is low, and the practical large-scale popularization is not facilitated.
Therefore, how to fully utilize the computing resources of the front-end and back-end devices, so as to improve the comprehensive resource utilization rate of the monitoring system, is a problem to be solved by those skilled in the art.
Disclosure of Invention
The application provides a processing method of monitoring service, which is used for solving the problem that the computing resources of front-end equipment and back-end equipment cannot be fully utilized in the prior art, so that the comprehensive resource utilization rate of a monitoring system is low. The method is applied to a monitoring system comprising front-end equipment and a server, and at least comprises the following steps:
the server receives a first processing result of the specified service, which is reported by the front-end equipment, and judges whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value or not through a preset confidence coefficient evaluation model;
if the judgment result is yes, the server takes the first processing result as the processing result of the appointed service;
if the judgment result is negative, the server processes the appointed service, acquires a second processing result of the appointed service, and takes the second processing result as the processing result of the appointed service.
Preferably, the first processing result is obtained after the front-end device processes the specified service through a preset detection algorithm, and after the judging result of the server is no, the method further includes:
acquiring the frequency that the confidence coefficient of the first processing result of the appointed service is lower than a preset confidence coefficient threshold value in a preset time interval of the front-end equipment;
if the frequency is greater than a preset frequency threshold, adjusting scene parameters of the detection algorithm according to the current scene of the front-end equipment;
and if the frequency is not greater than a preset frequency threshold, keeping the scene parameters of the detection algorithm unchanged.
Preferably, the server adjusts the scene parameters of the detection algorithm according to the current scene of the front-end device, specifically:
receiving an image sent by the front-end equipment and acquiring image information of the image;
determining a scene where the front-end equipment is currently located according to the image information of the image;
determining characteristic scene parameters corresponding to a scene where the front-end equipment is currently located according to a preset corresponding relation, wherein the corresponding relation is used for indicating the relation between the scene where the front-end equipment is located and the scene parameters of the detection algorithm;
and sending the characteristic scene parameters to the front-end equipment so that the front-end equipment adjusts the scene parameters of the detection algorithm into the characteristic scene parameters.
Preferably, the confidence evaluation model is:
wherein confidence is confidence, a i Weights for the i th sub-service of the specified service, c i And the confidence of the ith sub-service of the appointed service.
Preferably, the type of the specified service at least includes: target detection, target tracking, target recognition classification, and target feature refinement recognition.
Correspondingly, the application provides a server which is applied to a monitoring system comprising front-end equipment and the server, wherein the server at least comprises:
the judging module is used for receiving a first processing result of the specified service, which is reported by the front-end equipment, and judging whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value or not through a preset confidence coefficient evaluation model;
the first processing module is used for taking the first processing result as the processing result of the appointed service when the judgment result is yes;
and the second processing module is used for processing the appointed service when the judgment result is negative, acquiring a second processing result of the appointed service, and taking the second processing result as the processing result of the appointed service.
Preferably, the first processing result is obtained after the front-end device processes the specified service through a preset detection algorithm, and the server further includes:
the acquisition module is used for acquiring the frequency that the confidence coefficient of the first processing result of the appointed service is lower than a preset confidence coefficient threshold value in a preset time interval of the front-end equipment;
the adjusting module is used for adjusting scene parameters of the detection algorithm according to the current scene of the front-end equipment when the frequency is larger than a preset frequency threshold;
and the maintaining module is used for maintaining the scene parameters of the detection algorithm unchanged when the frequency is not greater than a preset frequency threshold.
Preferably, the adjusting module is specifically configured to:
receiving an image sent by the front-end equipment and acquiring image information of the image;
determining a scene where the front-end equipment is currently located according to the image information of the image;
determining characteristic scene parameters corresponding to a scene where the front-end equipment is currently located according to a preset corresponding relation, wherein the corresponding relation is used for indicating the relation between the scene where the front-end equipment is located and the scene parameters of the detection algorithm;
and sending the characteristic scene parameters to the front-end equipment so that the front-end equipment adjusts the scene parameters of the detection algorithm into the characteristic scene parameters.
Preferably, the confidence evaluation model is:
wherein confidence is confidence, ai is the weight of the ith sub-service of the specified service, and ci is the confidence of the ith sub-service of the specified service.
Preferably, the type of the specified service at least includes: target detection, target tracking, target recognition classification, and target feature refinement recognition.
By applying the technical scheme of the application, the server receives the first processing result of the specified service reported by the front-end equipment and judges whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value or not through a preset confidence coefficient evaluation model; if the judgment result is yes, the server takes the first processing result as the processing result of the appointed service; if the judgment result is negative, the server processes the appointed service, a second processing result is obtained, and the second processing result is used as the processing result of the appointed service. The first processing result of the front-end equipment is evaluated through the server, when the confidence coefficient of the first processing result is higher, the first processing result is adopted, and when the confidence coefficient of the first processing result is lower, the server further processes the appointed service, so that the resources of the front-end equipment are fully utilized on the basis of ensuring the service processing accuracy, and the comprehensive resource utilization rate of the monitoring system is improved.
Drawings
In order to more clearly illustrate the technical solutions of the present application, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of a method for processing a monitoring service according to an embodiment of the present application;
FIG. 2 is a flow chart of a target recognition method according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of a server according to the present application.
Detailed Description
As described in the background art, there are two main processing manners in the existing monitoring service processing scheme, one of which depends on the front-end device to process the monitoring service, and the method has the disadvantages that the processing capability of the front-end device processing chip is limited, the algorithm that the computing resource of the front-end device can be deployed is limited, and the requirement of accurate detection is difficult to meet. Secondly, the monitoring service is processed by the server at the back end, and the method has the defects that a large number of servers are needed at the back end, and the cost performance of the scheme is low due to the high price of the server at the back end, so that the method is not beneficial to actual popularization in a large range. As can be seen, in the prior art, the solution that relies solely on the front-end device or the back-end server does not fully utilize the resources of the front-end and the back-end.
Therefore, the application provides a processing method of monitoring service, which is used for solving the problem that the computing resources of front-end equipment and back-end equipment cannot be fully utilized in the prior art, so that the comprehensive resource utilization rate of a monitoring system is low. The method comprises the steps that a server receives a first processing result of a specified service, which is reported by front-end equipment, and judges whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value or not through a preset confidence coefficient evaluation model; if the judgment result is yes, the server takes the first processing result as the processing result of the appointed service; if the judgment result is negative, the server processes the appointed service, a second processing result is obtained, and the second processing result is used as the processing result of the appointed service. The first processing result of the front-end equipment is evaluated through the server, when the confidence coefficient of the first processing result is higher, the first processing result is adopted, and when the confidence coefficient of the first processing result is lower, the server further processes the appointed service, so that the resources of the front-end equipment are fully utilized on the basis of ensuring the service processing accuracy, and the comprehensive resource utilization rate of the monitoring system is improved.
As shown in fig. 1, a flow chart of a method for processing a monitoring service according to the present application is provided, and it should be noted that the present application is applied to a monitoring system including a front-end device and a server, and the front-end device in the present application may be specifically a monitoring device such as a monitoring camera. The server is used as the back-end equipment, can acquire the monitoring information of the front-end equipment in real time, and can control and manage the front-end equipment. Specifically, the application at least comprises the following steps:
s101, the server receives a first processing result of the specified service, which is reported by the front-end equipment, and judges whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value through a preset confidence coefficient evaluation model.
With the development of front-end equipment chips, the front-end equipment chips and the implementation of special and rapid computation unit operators, such as simple Gaussian background modeling, simple target tracking, CNN networks and the like, are realized. It can be seen that the head-end equipment is fully capable of handling the monitoring traffic. However, due to limited chip resources of the front-end device, these operators cannot meet the algorithm application of richer parameters, and thus, there may be a case that the processing result is inaccurate.
Aiming at the current development condition of the front-end equipment chip, in the embodiment of the application, the front-end equipment is used for processing the appointed service. When a first processing result of the specified service sent by the front-end equipment is received, the server evaluates the first processing result and judges whether the first processing result of the front-end equipment is needed to be adopted according to the evaluation result. Specifically, the server judges whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value through a preset confidence coefficient evaluation model, and if the confidence coefficient of the first processing result is higher than the preset confidence coefficient threshold value, the front-end equipment is informed of the first processing result of the appointed service. And if the confidence coefficient of the first processing result is not higher than the preset confidence coefficient threshold value, the front-end equipment is not trusted to the first processing result of the appointed service.
By the processing method, the server evaluates the first processing result of the front-end equipment, and the first processing result of the front-end equipment is adopted when the confidence coefficient of the first processing result of the front-end equipment is high, so that the service processing resource of the front-end equipment is fully utilized on the basis of ensuring the service processing accuracy, and the comprehensive resource utilization rate of the monitoring system is improved.
In a preferred embodiment of the present application, the confidence evaluation model is specifically:
wherein confidence is confidence, a i Weights for the i th sub-service of the specified service, c i For the specified serviceConfidence of the ith sub-service.
The confidence coefficient of each sub-service of the appointed service is weighted and summed through the confidence coefficient model, the confidence coefficient of each sub-service is provided by the front-end equipment, and the weight of each sub-service is obtained by evaluating the processing result of the sub-service.
It should be noted that the confidence model mentioned above is only a preferred model provided by the present application, and based on the core idea of the present application, other confidence models can be adopted by those skilled in the art, which does not affect the protection scope of the present application.
In a preferred embodiment of the present application, specifying the type of service includes at least: target detection, target tracking, target recognition classification, and target feature refinement recognition.
It should be noted that, the application scope of the method of the present application is not limited to processing the above disclosed specified service types, and based on the core concept of the present application, those skilled in the art may also process other types of specified services through the method provided by the present application, which does not affect the protection scope of the present application.
S102, if the judgment result is yes, the server takes the first processing result of the front-end equipment as the processing result of the appointed service.
In the embodiment of the application, if the confidence of the first processing result of the front-end equipment is higher than the preset confidence threshold, the first processing result of the front-end equipment is trusted. At this time, the server provides the first processing result with the front-end device, i.e., it is provided to the client as a processing result of the specified service.
It should be noted that, the confidence threshold may be set according to the requirement of the user. The higher the confidence threshold value is, the higher the accuracy requirement on the service processing result is; the lower the confidence threshold, the lower the accuracy requirement for the business process results. Therefore, the user can set the opposite credibility threshold according to the requirement of the user on the accuracy of the business processing result.
By the method, when the confidence coefficient of the first processing result of the front-end equipment is higher than the preset confidence coefficient threshold value, the processing result of the front-end equipment for the appointed service is adopted, and the server does not need to process the appointed service at the moment, so that the processing load of the server is reduced on the basis of ensuring the accuracy of service processing. Meanwhile, the method can fully utilize the service processing resources of the front-end equipment, and improves the comprehensive resource utilization rate of the monitoring system.
S103, if the judgment result is negative, the server processes the appointed service, acquires a second processing result of the appointed service, and takes the second processing result as the processing result of the appointed service.
In the embodiment of the application, if the confidence of the first processing result of the front-end equipment is not higher than the preset confidence threshold, the front-end equipment is not trusted to the first processing result of the designated service. At this time, the server processes the specified service, obtains a second processing result of the specified service, and provides the second processing result as a processing result of the specified service to the user.
When the confidence of the first processing result of the front-end equipment is low, that is, the first processing result of the front-end equipment is not trusted, the server further processes the specified service to ensure the accuracy of the processing result of the specified service.
In a preferred embodiment of the present application, the first processing result is obtained after the front-end device processes the specified service through a preset detection algorithm.
It should be noted that, after the front-end device processes the specified service through a preset detection algorithm, a first processing result of the specified service is obtained. When the front-end equipment is in different scenes, the scene parameters of the detection algorithm are different. Thus, when the confidence of the first processing result of the front-end device decreases, it may be caused by that the scene parameter of the detection algorithm does not match the scene in which the front-end device is currently located. Therefore, when the confidence of the first processing result of the front-end device decreases, the scene parameters of the detection algorithm need to be adjusted.
In a preferred embodiment of the present application, the adjustment of the scene parameters of the detection algorithm can be achieved by the following preferred scheme. Specifically, the scheme comprises the following steps:
(1) And acquiring the frequency that the confidence coefficient of the first processing result of the appointed service is lower than a preset confidence coefficient threshold value in a preset time interval of the front-end equipment.
Firstly, acquiring the frequency that the confidence coefficient of a first processing result of a specified service is lower than a preset confidence coefficient threshold value in a preset time interval (a preset period of time) of front-end equipment.
If the acquired frequency is greater than the preset frequency threshold, it is indicated that the scene where the front-end equipment is located may change, so that the scene parameters of the detection algorithm need to be adjusted.
If the acquired frequency is not greater than the preset frequency threshold, the scene where the front-end equipment is located is not changed, and the confidence is too low, which is possibly caused by accidental factors, so that the scene parameters of the detection algorithm do not need to be adjusted.
(2) If the acquired frequency is greater than a preset frequency threshold, adjusting scene parameters of the detection algorithm according to the current scene of the front-end equipment.
(3) If the acquired frequency is not greater than a preset frequency threshold, keeping the scene parameters of the detection algorithm unchanged.
In a preferred embodiment of the present application, the above adjustment of the scene parameters of the detection algorithm may be specifically implemented by the following preferred scheme, which specifically includes the following steps:
(1) And receiving the image sent by the front-end equipment and acquiring the image information of the image.
Firstly, analyzing an image sent by front-end equipment to acquire image information in the image. The image information at least comprises statistical information such as image contrast, target contrast, image histogram, target histogram, scene brightness, color, texture, front and rear frame difference image data and the like.
(2) And determining the current scene of the front-end equipment according to the image information of the image.
And analyzing the image information of the front-end equipment image to determine the current scene of the front-end equipment. After determining the current scene of the front-end equipment, further determining parameters required to be set for the front-end equipment.
(3) And determining characteristic scene parameters corresponding to the scene where the front-end equipment is currently located according to the preset corresponding relation. The corresponding relation is used for indicating the relation between the scene where the front-end equipment is located and the scene parameters of the detection algorithm.
The corresponding relation between the scene where the front-end equipment is located and the scene parameters of the detection algorithm is stored in the server in advance. After the scene of the front-end equipment is acquired, the characteristic scene parameters which are required to be set currently by the corresponding relation detection algorithm are detected.
(4) And generating the acquired characteristic scene parameters to the front-end equipment so that the front-end equipment can adjust the scene parameters of the detection algorithm into the characteristic scene parameters.
Through the above preferred embodiment, when the confidence of the first processing result of the front-end device is low, the server adjusts the scene parameters of the detection algorithm in the front-end device according to the current scene of the front-end device, so as to improve the confidence of the detection result of the front-end device.
As can be seen from the description of the above embodiment, the server receives the first processing result of the specified service reported by the front-end device, and determines, through a preset confidence evaluation model, whether the confidence of the first processing result is higher than a preset confidence threshold; if the judgment result is yes, the server takes the first processing result as the processing result of the appointed service; if the judgment result is negative, the server processes the appointed service, a second processing result is obtained, and the second processing result is used as the processing result of the appointed service. The first processing result of the front-end equipment is evaluated through the server, when the confidence coefficient of the first processing result is higher, the first processing result is adopted, and when the confidence coefficient of the first processing result is lower, the server further processes the appointed service, so that the resources of the front-end equipment are fully utilized on the basis of ensuring the service processing accuracy, and the comprehensive resource utilization rate of the monitoring system is improved.
In order to further explain the technical idea of the present application, the technical scheme of the present application will now be described with reference to a specific implementation flow.
The application provides a system scheme for realizing intelligent analysis algorithm by combining computing resources of front end chips and back end chips and carrying out algorithm coordinated allocation. The application is applied to monitoring business and at least comprises the following steps: target detection, target tracking, target identification classification, feature refinement identification and the like. And constructing a confidence evaluation model aiming at the algorithm effect of each module, simultaneously detecting the scene characteristics of the current video image in real time, and transmitting data to a back-end server through a high-speed bus. The back-end server decides whether the front-end result is credible or not according to the confidence coefficient mode, and the video is reprocessed in an unreliable mode until the target analysis algorithm obtains the result. The decision information comprises resource information of a front-end system and algorithm service. The technical effects of the application include: the computing advantage of the front-end IPC (monitoring camera) is fully exerted, the redundant computing capacity and memory resources and the front-end image which is not damaged by video compression are utilized, the powerful computing resources and storage resources of the back-end server equipment are combined on the basis, the algorithm accuracy is improved, and the system resource requirement is reduced.
Taking intelligent target recognition as an example, as shown in fig. 2, a specific implementation process includes the following steps:
s201, initializing the system, and starting the front-end and back-end systems.
S202, monitoring and collecting system start scene information in real time. Including but not limited to image contrast, histogram, average brightness, shutter, aperture, ISO, etc., scene related parameters.
S203, video scene moving object detection, taking an object detection classifier as an example, but not limited to, the algorithm. Pedestrian detection is detected by adopting an adboost classifier. The classifier detection output includes the number, location, size, and confidence of the target detection.
S204, the rear-end server acquires front-end target detection information and confidence, establishes a confidence decision model, and evaluates the detection result of the front-end device. Judging front-end equipmentWhether the detection result of (2) is higher than a preset threshold value theta 1 . If yes, go to S205, otherwise go to S206.
S205, adopting a target detection result of the front-end equipment, and directly adopting the number, the position and the size of targets by a subsequent module.
S206, the back end establishes an evaluation model according to the real-time information of the scene environment. Evaluation includes, but is not limited to: and outputting detection results by correspondingly selecting back-end target detection algorithms according to scene factors such as day, night, shadow, rainy days and the like.
S207, obtaining that the confidence coefficient of the detection result of the front-end equipment is lower than a threshold value theta 1 Is a frequency of (2).
S208, judging that the confidence of the detection result of the front-end equipment is lower than the threshold value theta 1 If the frequency of (2) is higher than the preset frequency threshold, if not, go to S209, otherwise go to S210.
S209, keeping the front-end algorithm parameters unchanged.
S210, selecting parameters of a front-end algorithm according to the scene, and sending the updated parameters to the front end.
S211, the front end updates the target detection parameters and enters the next target detection cycle. Until the system exits.
The application also establishes a confidence evaluation model, which comprises the following steps:
(1) Front-end video scene detection evaluation, statistics of front-end image information, including but not limited to image contrast, target contrast, image histogram, target histogram, scene brightness, color, texture, front-to-back frame difference map data, etc.
(2) Environmental factor prediction, determining a scene from the front-end image and sensor data information, including but not limited to: scenes such as day, night, rainy days, sunny days, cloudy days and the like.
(3) The front end target detection algorithm module outputs results and confidence, and the front end detection results comprise: target number, target ID, target position, speed, etc.
(4) And the posterior statistical model of the algorithm accuracy on the scene and the parameters is used as the algorithm confidence evaluation model data.
The specific evaluation model is as follows:
wherein confidence is confidence, a i Weights for the i th sub-service of the specified service, c i And the confidence of the ith sub-service of the appointed service.
According to the description of the specific embodiment, the server receives the first processing result of the specified service, which is reported by the front-end equipment, and judges whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value or not through a preset confidence coefficient evaluation model; if the judgment result is yes, the server takes the first processing result as the processing result of the appointed service; if the judgment result is negative, the server processes the appointed service, a second processing result is obtained, and the second processing result is used as the processing result of the appointed service. The first processing result of the front-end equipment is evaluated through the server, when the confidence coefficient of the first processing result is higher, the first processing result is adopted, and when the confidence coefficient of the first processing result is lower, the server further processes the appointed service, so that the resources of the front-end equipment are fully utilized on the basis of ensuring the service processing accuracy, and the comprehensive resource utilization rate of the monitoring system is improved.
In order to achieve the above technical objective, as shown in fig. 3, the present application proposes a server applied to a monitoring system including a front-end device and the server, where the server at least includes:
the judging module 301 receives a first processing result of the specified service, which is reported by the front-end equipment, and judges whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value through a preset confidence coefficient evaluation model;
the first processing module 302 takes the first processing result as the processing result of the specified service when the judging result is yes;
and the second processing module 303 processes the specified service to obtain a second processing result of the specified service when the judging result is negative, and takes the second processing result as the processing result of the specified service.
Preferably, the first processing result is obtained after the front-end device processes the specified service through a preset detection algorithm, and the server further includes:
the acquisition module is used for acquiring the frequency that the confidence coefficient of the first processing result of the appointed service is lower than a preset confidence coefficient threshold value in a preset time interval of the front-end equipment;
the adjusting module is used for adjusting scene parameters of the detection algorithm according to the current scene of the front-end equipment when the frequency is larger than a preset frequency threshold;
and the maintaining module is used for maintaining the scene parameters of the detection algorithm unchanged when the frequency is not greater than a preset frequency threshold.
Preferably, the adjusting module is specifically configured to:
receiving an image sent by the front-end equipment and acquiring image information of the image;
determining a scene where the front-end equipment is currently located according to the image information of the image;
determining characteristic scene parameters corresponding to a scene where the front-end equipment is currently located according to a preset corresponding relation, wherein the corresponding relation is used for indicating the relation between the scene where the front-end equipment is located and the scene parameters of the detection algorithm;
and sending the characteristic scene parameters to the front-end equipment so that the front-end equipment adjusts the scene parameters of the detection algorithm into the characteristic scene parameters.
Preferably, the confidence evaluation model is:
wherein confidence is confidence, ai is the weight of the ith sub-service of the specified service, and ci is the confidence of the ith sub-service of the specified service.
Preferably, the type of the specified service at least includes: target detection, target tracking, target recognition classification, and target feature refinement recognition.
According to the description of the specific equipment, the server receives the first processing result of the specified service, which is reported by the front-end equipment, and judges whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value or not through a preset confidence coefficient evaluation model; if the judgment result is yes, the server takes the first processing result as the processing result of the appointed service; if the judgment result is negative, the server processes the appointed service, a second processing result is obtained, and the second processing result is used as the processing result of the appointed service. The first processing result of the front-end equipment is evaluated through the server, when the confidence coefficient of the first processing result is higher, the first processing result is adopted, and when the confidence coefficient of the first processing result is lower, the server further processes the appointed service, so that the resources of the front-end equipment are fully utilized on the basis of ensuring the service processing accuracy, and the comprehensive resource utilization rate of the monitoring system is improved.
The last explanation is: the above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art will appreciate that; the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the corresponding technical solutions, which are defined by the scope of the appended claims.
From the above description of the embodiments, it will be clear to those skilled in the art that the present application may be implemented in hardware, or may be implemented by means of software plus necessary general hardware platforms. Based on such understanding, the technical solution of the present application may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (may be a CD-ROM, a U-disk, a mobile hard disk, etc.), and includes several instructions for causing a computer device (may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective implementation scenario of the present application.
Those skilled in the art will appreciate that the drawing is merely a schematic illustration of a preferred implementation scenario and that the modules or flows in the drawing are not necessarily required to practice the application.
Those skilled in the art will appreciate that modules in an apparatus in an implementation scenario may be distributed in an apparatus in an implementation scenario according to an implementation scenario description, or that corresponding changes may be located in one or more apparatuses different from the implementation scenario. The modules of the implementation scenario may be combined into one module, or may be further split into a plurality of sub-modules.
The above-mentioned inventive sequence numbers are merely for description and do not represent advantages or disadvantages of the implementation scenario.
The foregoing disclosure is merely illustrative of some embodiments of the application, and the application is not limited thereto, as modifications may be made by those skilled in the art without departing from the scope of the application.
Claims (8)
1. The processing method of the monitoring service is characterized by being applied to a monitoring system comprising front-end equipment and a server, and the method at least comprises the following steps:
the server receives a first processing result of the specified service, which is reported by the front-end equipment, and judges whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value or not through a preset confidence coefficient evaluation model;
if the judgment result is yes, the server takes the first processing result as the processing result of the appointed service;
if the judgment result is negative, the server processes the appointed service, acquires a second processing result of the appointed service, and takes the second processing result as the processing result of the appointed service;
the first processing result is obtained after the front-end equipment processes the specified service through a preset detection algorithm, and after the judging result of the server is negative, the method further comprises the following steps:
acquiring the frequency that the confidence coefficient of the first processing result of the appointed service is lower than a preset confidence coefficient threshold value in a preset time interval of the front-end equipment;
if the frequency is greater than a preset frequency threshold, adjusting scene parameters of the detection algorithm according to the current scene of the front-end equipment;
the server adjusts scene parameters of the detection algorithm according to the current scene of the front-end equipment, specifically:
receiving an image sent by the front-end equipment and acquiring image information of the image;
determining a scene where the front-end equipment is currently located according to the image information of the image;
determining characteristic scene parameters corresponding to a scene where the front-end equipment is currently located according to a preset corresponding relation, wherein the corresponding relation is used for indicating the relation between the scene where the front-end equipment is located and the scene parameters of the detection algorithm;
and sending the characteristic scene parameters to the front-end equipment so that the front-end equipment adjusts the scene parameters of the detection algorithm into the characteristic scene parameters.
2. The method of claim 1, wherein,
and if the frequency is not greater than a preset frequency threshold, keeping the scene parameters of the detection algorithm unchanged.
3. The method of claim 1, wherein,
the confidence evaluation model is as follows:
wherein confidence is confidence, a i Weights for the i th sub-service of the specified service, c i And the confidence of the ith sub-service of the appointed service.
4. The method of any one of claim 1 to 3, wherein,
the specified service types at least comprise: target detection, target tracking, target recognition classification, and target feature refinement recognition.
5. A server for use in a monitoring system comprising a head-end and said server, said server comprising at least:
the judging module is used for receiving a first processing result of the specified service, which is reported by the front-end equipment, and judging whether the confidence coefficient of the first processing result is higher than a preset confidence coefficient threshold value or not through a preset confidence coefficient evaluation model;
the first processing module is used for taking the first processing result as the processing result of the appointed service when the judgment result is yes;
the second processing module is used for processing the appointed service when the judgment result is negative, acquiring a second processing result of the appointed service, and taking the second processing result as the processing result of the appointed service;
the first processing result is obtained after the front-end device processes the specified service through a preset detection algorithm, and the server further comprises:
the acquisition module is used for acquiring the frequency that the confidence coefficient of the first processing result of the appointed service is lower than a preset confidence coefficient threshold value in a preset time interval of the front-end equipment;
the adjusting module is used for adjusting scene parameters of the detection algorithm according to the current scene of the front-end equipment when the frequency is larger than a preset frequency threshold;
the adjusting module is specifically used for:
receiving an image sent by the front-end equipment and acquiring image information of the image;
determining a scene where the front-end equipment is currently located according to the image information of the image;
determining characteristic scene parameters corresponding to a scene where the front-end equipment is currently located according to a preset corresponding relation, wherein the corresponding relation is used for indicating the relation between the scene where the front-end equipment is located and the scene parameters of the detection algorithm;
and sending the characteristic scene parameters to the front-end equipment so that the front-end equipment adjusts the scene parameters of the detection algorithm into the characteristic scene parameters.
6. The server of claim 5, wherein the server further comprises:
and the maintaining module is used for maintaining the scene parameters of the detection algorithm unchanged when the frequency is not greater than a preset frequency threshold.
7. The server according to claim 5, wherein,
the confidence evaluation model is as follows:
wherein confidence is confidence, a i Weights for the i th sub-service of the specified service, c i And the confidence of the ith sub-service of the appointed service.
8. The server according to any one of claim 5 to 7,
the specified service types at least comprise: target detection, target tracking, target recognition classification, and target feature refinement recognition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611117138.5A CN106682590B (en) | 2016-12-07 | 2016-12-07 | Processing method of monitoring service and server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611117138.5A CN106682590B (en) | 2016-12-07 | 2016-12-07 | Processing method of monitoring service and server |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106682590A CN106682590A (en) | 2017-05-17 |
| CN106682590B true CN106682590B (en) | 2023-08-22 |
Family
ID=58868770
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611117138.5A Active CN106682590B (en) | 2016-12-07 | 2016-12-07 | Processing method of monitoring service and server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106682590B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110557633B (en) * | 2019-08-28 | 2021-06-29 | 深圳大学 | Compression transmission method, system and computer readable storage medium for image data |
| CN111144495B (en) * | 2019-12-27 | 2024-03-22 | 浙江宇视科技有限公司 | Service distribution method, device and medium |
| CN112966608A (en) * | 2021-03-05 | 2021-06-15 | 哈尔滨工业大学 | Target detection method, system and storage medium based on edge-side cooperation |
| CN117177203B (en) * | 2023-09-18 | 2024-06-07 | 深圳市哈烁实业有限公司 | Intelligent park building security method based on wireless transmission |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1629875A (en) * | 2003-12-15 | 2005-06-22 | 中国科学院自动化研究所 | Distributed human face detecting and identifying method under mobile computing environment |
| CN102158689A (en) * | 2011-05-17 | 2011-08-17 | 无锡中星微电子有限公司 | Video monitoring system and method |
| CN102300287A (en) * | 2011-08-12 | 2011-12-28 | 盛乐信息技术（上海）有限公司 | Wireless hotspot connection method and apparatus thereof |
| WO2012142797A1 (en) * | 2011-04-18 | 2012-10-26 | 中兴通讯股份有限公司 | Video monitoring system and method |
| CN103208014A (en) * | 2012-01-13 | 2013-07-17 | 施亿民 | Image recognition system and operation method thereof |
| CN103297512A (en) * | 2013-05-15 | 2013-09-11 | 吴玉平 | System and method for intelligent face recognition area network monitoring based on cloud technology |
| CN103434484A (en) * | 2013-08-20 | 2013-12-11 | 安科智慧城市技术（中国）有限公司 | Vehicle-mounted identification and authentication device, mobile terminal and intelligent vehicle key control system and method |
| CN104159071A (en) * | 2014-07-11 | 2014-11-19 | 深圳瞭望通达科技有限公司 | Intelligent target identification device, system and method based on cloud service |
| CN104580338A (en) * | 2013-10-29 | 2015-04-29 | 华为技术有限公司 | Service processing method, system and equipment |
| CN104575503A (en) * | 2015-01-16 | 2015-04-29 | 广东美的制冷设备有限公司 | Speech recognition method and device |
| CN104954480A (en) * | 2015-06-24 | 2015-09-30 | 深圳市兰丁科技有限公司 | Equipment parameter setting method and device |
| CN105095919A (en) * | 2015-09-08 | 2015-11-25 | 北京百度网讯科技有限公司 | Image recognition method and image recognition device |
| CN105657067A (en) * | 2016-03-24 | 2016-06-08 | 网易（杭州）网络有限公司 | Game verification method and device, game server and verification server |
-
2016
- 2016-12-07 CN CN201611117138.5A patent/CN106682590B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1629875A (en) * | 2003-12-15 | 2005-06-22 | 中国科学院自动化研究所 | Distributed human face detecting and identifying method under mobile computing environment |
| WO2012142797A1 (en) * | 2011-04-18 | 2012-10-26 | 中兴通讯股份有限公司 | Video monitoring system and method |
| CN102158689A (en) * | 2011-05-17 | 2011-08-17 | 无锡中星微电子有限公司 | Video monitoring system and method |
| CN102300287A (en) * | 2011-08-12 | 2011-12-28 | 盛乐信息技术（上海）有限公司 | Wireless hotspot connection method and apparatus thereof |
| CN103208014A (en) * | 2012-01-13 | 2013-07-17 | 施亿民 | Image recognition system and operation method thereof |
| CN103297512A (en) * | 2013-05-15 | 2013-09-11 | 吴玉平 | System and method for intelligent face recognition area network monitoring based on cloud technology |
| CN103434484A (en) * | 2013-08-20 | 2013-12-11 | 安科智慧城市技术（中国）有限公司 | Vehicle-mounted identification and authentication device, mobile terminal and intelligent vehicle key control system and method |
| CN104580338A (en) * | 2013-10-29 | 2015-04-29 | 华为技术有限公司 | Service processing method, system and equipment |
| CN104159071A (en) * | 2014-07-11 | 2014-11-19 | 深圳瞭望通达科技有限公司 | Intelligent target identification device, system and method based on cloud service |
| CN104575503A (en) * | 2015-01-16 | 2015-04-29 | 广东美的制冷设备有限公司 | Speech recognition method and device |
| CN104954480A (en) * | 2015-06-24 | 2015-09-30 | 深圳市兰丁科技有限公司 | Equipment parameter setting method and device |
| CN105095919A (en) * | 2015-09-08 | 2015-11-25 | 北京百度网讯科技有限公司 | Image recognition method and image recognition device |
| CN105657067A (en) * | 2016-03-24 | 2016-06-08 | 网易（杭州）网络有限公司 | Game verification method and device, game server and verification server |
Non-Patent Citations (1)
| Title |
|---|
| 网游客户端计算服务器验证;金庆;《https://blog.csdn.net/jq0123/article/details/5942543》;20101015;第1页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106682590A (en) | 2017-05-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| del Postigo et al. | Vacant parking area estimation through background subtraction and transience map analysis | |
| US9460361B2 (en) | Foreground analysis based on tracking information | |
| CN106682590B (en) | Processing method of monitoring service and server | |
| US9159137B2 (en) | Probabilistic neural network based moving object detection method and an apparatus using the same | |
| KR102170992B1 (en) | Multi-functional vehicle number recognition system, and method thereof | |
| Khalifa et al. | [Retracted] Vehicle Detection for Vision‐Based Intelligent Transportation Systems Using Convolutional Neural Network Algorithm | |
| CN111325051B (en) | Face recognition method and device based on face image ROI selection | |
| CN111462155B (en) | Motion detection method, device, computer equipment and storage medium | |
| KR102122850B1 (en) | Solution for analysis road and recognition vehicle license plate employing deep-learning | |
| Ali et al. | Autonomous road surveillance system: A proposed model for vehicle detection and traffic signal control | |
| CN110796580B (en) | Intelligent traffic system management method and related products | |
| Abidin et al. | A systematic review of machine-vision-based smart parking systems | |
| KR102584708B1 (en) | System and Method for Crowd Risk Management by Supporting Under and Over Crowded Environments | |
| CN105608464A (en) | Face recognition mobile terminal solving system and solving method | |
| Magrini et al. | Computer vision on embedded sensors for traffic flow monitoring | |
| CN109800675A (en) | A kind of method and device of the identification image of determining face object | |
| CN111929672A (en) | Method and device for determining movement track, storage medium and electronic device | |
| KR20170095599A (en) | System and method for video searching | |
| KR102421043B1 (en) | Apparatus for Processing Images and Driving Method Thereof | |
| CN111179610A (en) | Control method and device of traffic signal equipment | |
| CN111064924A (en) | Video monitoring method and system based on artificial intelligence | |
| Vujović et al. | Traffic video surveillance in different weather conditions | |
| Liu et al. | Research on Detection Algorithm of Wheel Position based on CenterNet | |
| CN113963310A (en) | People flow detection method and device for bus station and electronic equipment | |
| Sankaranarayanan et al. | Improved Vehicle Detection Accuracy and Processing Time for Video Based ITS Applications |
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 |