CN111353338A - Energy efficiency improvement method based on business hall video monitoring - Google Patents
Energy efficiency improvement method based on business hall video monitoring Download PDFInfo
- Publication number
- CN111353338A CN111353338A CN201811573211.9A CN201811573211A CN111353338A CN 111353338 A CN111353338 A CN 111353338A CN 201811573211 A CN201811573211 A CN 201811573211A CN 111353338 A CN111353338 A CN 111353338A
- Authority
- CN
- China
- Prior art keywords
- business
- time
- handling
- video
- image
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000012544 monitoring process Methods 0.000 title claims abstract description 40
- 230000006872 improvement Effects 0.000 title claims abstract description 9
- 238000009826 distribution Methods 0.000 claims description 26
- 238000001514 detection method Methods 0.000 claims description 13
- 230000033001 locomotion Effects 0.000 claims description 10
- 230000002688 persistence Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 description 18
- 230000006399 behavior Effects 0.000 description 17
- 238000004458 analytical method Methods 0.000 description 15
- 230000002159 abnormal effect Effects 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 10
- 230000008569 process Effects 0.000 description 10
- 238000000605 extraction Methods 0.000 description 8
- 238000013528 artificial neural network Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 7
- 238000007689 inspection Methods 0.000 description 7
- 230000009471 action Effects 0.000 description 6
- 239000000203 mixture Substances 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000013518 transcription Methods 0.000 description 5
- 230000035897 transcription Effects 0.000 description 5
- RWSOTUBLDIXVET-UHFFFAOYSA-N Dihydrogen sulfide Chemical compound S RWSOTUBLDIXVET-UHFFFAOYSA-N 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000003068 static effect Effects 0.000 description 3
- 206010000117 Abnormal behaviour Diseases 0.000 description 2
- 238000013499 data model Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 230000036544 posture Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000002996 emotional effect Effects 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 210000000697 sensory organ Anatomy 0.000 description 1
- 230000005236 sound signal Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- 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
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
- G06V20/42—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items of sport video content
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention provides an energy efficiency improvement method based on business hall video monitoring, which comprises the following steps: A. acquiring real-time video monitoring and confirming personnel information in the video monitoring; B. confirming the handling time of different services based on the personnel information so as to determine the handling efficiency of different employees; C. adjusting staff allocation based on the transaction efficiency; the personnel information includes at least identification, location and tracking of customers and employees. Therefore, the time for the client to handle the business and the efficiency for the staff to handle the business can be judged by identifying, positioning and tracking the client and the staff, so that the staff allocation is adjusted to achieve the highest operation efficiency of the business hall.
Description
Technical Field
The invention relates to the technical field of video monitoring, in particular to an energy efficiency improvement method based on business hall video monitoring.
Background
The prior art has a plurality of difficulties for monitoring the business hall, such as: the appearance is very different, for example at different angles, the appearance of the pedestrian is very different. Pedestrians in different postures and different clothes (including attachments such as umbrella holding, hat wearing, scarf wearing, luggage carrying and the like) have great appearance difference. For another example, the tracking is difficult, and due to the fact that pedestrians are very dense and have serious shielding, only a part of a human body can be seen. For another example, the background is complex, the background generally faced by pedestrian detection is very complex, and the appearance, shape, color and texture of some objects are very similar to those of human bodies, so that the algorithms cannot accurately distinguish the objects.
The supervision for the operation of the business hall usually depends on the video monitoring technology of the business hall, and if the business hall cannot be effectively supervised according to the video monitoring technology, an elbow exists for improving the energy efficiency of the business hall.
Disclosure of Invention
The invention mainly aims to provide an energy efficiency improvement method based on business hall video monitoring, which comprises the following steps:
A. acquiring real-time video monitoring and confirming personnel information in the video monitoring;
B. confirming the handling time of different services based on the personnel information so as to determine the handling efficiency of different employees;
C. adjusting staff allocation based on the transaction efficiency;
the personnel information includes at least identification, location and tracking of customers and employees.
Therefore, the time for the client to handle the business and the efficiency for the staff to handle the business can be judged by identifying, positioning and tracking the client and the staff, so that the staff allocation is adjusted to achieve the highest operation efficiency of the business hall.
In step a, the confirming of the person information in the video monitoring includes:
performing background modeling on the video monitoring image to determine a background model and a foreground image;
and carrying out steps including human shape detection, human face recognition, personnel positioning and personnel statistics in the foreground image.
Therefore, the identification, positioning and tracking of the client and the staff are realized.
And step A, performing voice recognition on the video monitoring.
Therefore, aiming at the condition that very small actions such as eating, telephone chatting and the like have identification difficulty in the video, the voice recognition mode can be matched with video monitoring to realize the video monitoring. Therefore, the video monitoring is more effectively carried out on the business hall.
Wherein the step of performing background modeling on the video-monitored image comprises: and establishing a mixed Gaussian model for the color value of each pixel point in the image, and determining a real background according to the difference of the persistence and the variability of each Gaussian distribution through sampling observation for a period of time, wherein the Gaussian distribution corresponding to the real background is used as a background model.
Therefore, the background modeling realized by the Gaussian mixture model is used for describing a multi-modal scene, and pixel points belonging to a target are directly detected in a new frame of image while the background model is continuously updated, so that links of calculating a difference image, a binary image and the like are reduced, and the detection speed is effectively improved.
Wherein the step B comprises:
the method comprises the steps of identifying, positioning and tracking the client, obtaining the time of the client entering a business hall, and obtaining the waiting time and the handling time of the user for handling the business according to the time of staying in different business areas, thereby counting the business handling efficiency.
The step B further comprises the following steps: and comparing the movement track or the behavior pattern of the client with the normal movement track or the normal behavior pattern, thereby monitoring the safety behavior of the client.
Therefore, the service handling behavior and the safety behavior of the client can be effectively detected and judged.
Wherein, the step B further comprises: and identifying and positioning each worker, and analyzing the leaving time according to the time node of each worker leaving and returning to the station, thereby counting the business handling efficiency of each worker.
Therefore, the business handling efficiency of the staff can be counted and evaluated.
Step C, distributing specific staff to different clients for handling specific services based on the historical data of the handling efficiency;
the particular employee includes the most efficient employee to handle the particular business.
Therefore, based on the historical frequency of the clients entering and exiting the business hall, the business types and the number of the clients are matched, the business points concerned by the staff are analyzed, and relevant suggestions for providing personalized and precise services for the staff are provided. For example, if a client has a long time to transact a certain service, the shortest employee is scheduled to receive the service, thereby improving efficiency.
Drawings
Fig. 1 is a schematic diagram of a first embodiment of an energy efficiency improvement method based on business hall video monitoring;
fig. 2 is a schematic diagram of a second embodiment of an energy efficiency improvement method based on business hall video monitoring.
Detailed Description
This application is through video analysis technique, and several aspects such as the process is handled to business office operational environment, staff, customer's business carry out the degree of depth and excavate, and supplementary managers carries out manual intervention to the efficiency that business office business was handled, if increase the business and handle the seat quantity, carries out artifical reposition of redundant personnel and business guide to the customer to improve the efficiency that the business was handled to the customer. The efficiency of the business hall for receiving the customer is evaluated through the statistical analysis of the time spent on the business handling of the customer. In addition, through analyzing and extracting the behaviors, tracks and the like of a large number of business hall clients handling business, a data model of normal behavior patterns, action tracks and waiting areas of the clients in the business hall is established, and the tracks and behaviors of the staff entering the business hall are compared through the data model to judge whether the behavior patterns are normal or not.
The video analysis technique mainly comprises: and (4) monitoring the customer experience and supervising the behaviors of service personnel in a business hall.
The customer experience monitoring comprises business hall efficiency monitoring analysis, customer trajectory analysis and abnormal behavior monitoring.
The business hall service personnel behavior supervision comprises the inspection of business hall service personnel going on duty and leaving duty.
FIG. 1 is a flow chart of customer experience monitoring, comprising the steps of:
s101: and (5) real-time video monitoring.
Taking a business hall as an example, the camera devices are erected at different positions of the business hall, so that the business hall is set as a monitoring area.
S102: the video processing process comprises the following substeps:
firstly, preprocessing such as denoising, defogging, deblurring and the like is carried out on the video image by using a sharpening algorithm.
Secondly, different strategies are adopted for different scenes: video action recognition is adopted for the behaviors of people (clients) in a business hall (such as running, fighting and the like). For some actions of business hall workers on the workstation, recognition based on static images is adopted. The static image-based recognition includes: when the static human face attendance machine performs the human face recognition, a person needs to continuously adjust the face angle in front of the attendance machine so as to be convenient for the attendance machine to snapshot the positive and qualified human face image for comparison. The algorithm of the device is relatively simple. This is relatively long. The video motion recognition is to judge, track, snapshot and compare the human face in the normal walking process of a person, needs to deal with the problems of light, angle, fuzzy snapshot and the like, and has quick and accurate comparison requirements.
However, it should be noted that there is technical difficulty in recognizing very small actions such as eating, telephone chatting, etc. under the far-lens video, so that the part can be realized by combining or speech recognition.
Modeling a video image background;
background modeling technology is a method for detecting moving objects in video images, and the basic idea is to model the background of an image, and the principle is to compare the current image with a background model and determine a foreground object (the moving object to be detected) according to the comparison result.
Background modeling using a gaussian mixture model is a method for describing a multi-modal scene. The method can directly detect the pixel points belonging to the target in a new frame of image while continuously updating the background model, thereby reducing links of calculating a differential image, a binary image and the like and effectively improving the detection speed. The basic idea of the algorithm is: and establishing a mixed Gaussian model for the color value of each pixel point in the image, and judging which distribution is closer to a real background according to the difference of the persistence and the variability of each Gaussian distribution through sampling observation for a period of time, wherein the Gaussian distribution is used as the background model. If the color value of the pixel point in the image does not conform to the Gaussian distribution, the pixel point is considered as the target point. The specific calculation method is as follows.
Setting the observation value of the pixel point with the position (x0, y0) in the image in a period of time as follows:
{X1，…,Xt}＝{I(x0,y0i) 1 ≦ i ≦ t } -equation 1
The observed value in formula 1 is modeled by using a plurality of Gaussian distributions, and the probability of the color value of the current pixel point is obtained as follows:
wherein K is the number of Gaussian distributions (usually 3-5); omegai,tThe estimated value of the weight is the probability that the pixel point belongs to the ith Gaussian distribution at the moment t; mu.si,tIs the mean value of the ith Gaussian distribution at the time t; sigmai,tη is a probability density function of the Gaussian distribution:
for simplicity of calculation, assuming that three components (R, G, B) of the color value of the pixel point are independent of each other and have the same variance, the covariance matrix in equation 3 can be written as:
thus, a Gaussian mixture model of the color values of the observed pixel points (x0, y0) is established. For a pixel point (x0, y0, t) in the input image, comparing the color value with the existing K Gaussian distributions, judging whether the pixel point is matched with the existing Gaussian distributions, and if the pixel point is matched with the existing Gaussian distributions, determining that the pixel point is a background point. By "matching", equation 5 is satisfied.
|(Xt-μi,t-1)|＜TH×σi,t-1-formula 5
Wherein, mui,t-1TH is the mean value of the ith Gaussian distribution at time t-1, and is usually 2.5, sigmai,t-1Is the ith Gaussian distributionStandard deviation at time t-1.
If no matched Gaussian distribution is found, the color value of the input pixel is taken as a mean value to establish a new Gaussian distribution, and the distribution with the largest variance and the lowest weight in the previous K Gaussian distributions is replaced, so that the background model is reestablished.
If there is a matching Gaussian distribution, the parameters in the background model are updated as follows:
ωk,t＝(1-α)ωk,t-1+α(Mk,t) -formula 6
μt＝(1-ρ)μt-1+ρXt-formula 7
Wherein α is model learning rate, 1/α represents model parameter change rate, and matched Gaussian distribution Mk,t1, the remaining unmatched Gaussian distribution Mk,tIs 0; the formulas 7 and 8 only aim at the matched Gaussian distribution, and the parameters corresponding to the rest unmatched Gaussian distributions are kept unchanged; ρ is a parameter learning rate, defined as:
ρ＝αη(Xt|μk,σk) -formula 9
In this way, the gaussian mixture model is updated by expressions 6 to 9. The method of the Gaussian mixture model can fully represent the multi-modal characteristics of the scene; the method can quickly adapt to the change of the background, and can detect the moving target even when the scene has illumination change and small-amplitude repeated motion; if an object enters the scene and stays for a long time to become a background, the Gaussian mixture model method can also update the background model in time.
And monitoring the moving target in the video image according to the background model established in the step, and determining all foreground targets which are different from the background model in the video image. The foreground object comprises a track of a client, and the video action recognition technology is adopted, and the method specifically comprises the following steps:
human shape detection:
in video surveillance, automatically searching for human bodies in a scene is considered as the primary preprocessing step for understanding human activities. The method adopts Histogram of Oriented Gradient (HOG) to describe an interested target, simultaneously adopts a hidden Support Vector Machine (LSVM) classifier to train and classify the Histogram of Oriented Gradient, and then uses the hidden Support Vector Machine classifier to detect the image. In the process, an interested target needs to be divided into different regions manually, and Scale-invariant feature transform (SIFT) of each region is calculated. In combination with these features, an iterative algorithm is used to train a classifier (e.g., the Adaboost algorithm) to detect the target. The method can well solve the problem that the shape of the target changes due to movement, is insensitive to illumination change and micro offset, and can well detect and label the humanoid target in the real scene video image.
Object tracking:
object tracking techniques refer to the extraction of foreground varying regions from a background image from a sequence of images. The method comprises the steps of feature extraction, feature matching, region-based tracking and model-based tracking.
a. Feature extraction
Feature extraction refers to extracting a rendering feature of an image from an original image of a scene.
The ideal image characteristics should have the following characteristics:
the characteristics should have intuitive significance and accord with the visual characteristics of people;
the characteristics should have better classification capability and can distinguish different image contents;
feature calculation should be relatively simple to facilitate rapid identification;
the features should have invariance to image translation, rotation, scale changes, etc.
The features of a moving object commonly used in object tracking mainly include color, texture, edges, block features, optical flow features, perimeter, area, centroid, corner points, and the like. And extracting effective characteristics insensitive to scale expansion, deformation and brightness change.
b. Feature matching
The feature matching is to match the features of the target between frames and track the target with the optimal matching. The tracking algorithm comprises tracking based on binarization target image matching, tracking based on edge feature matching or corner feature matching, tracking based on target gray scale feature matching, tracking based on target color feature matching and the like.
The tracking algorithm based on the characteristics is insensitive to the changes of the scale, the deformation, the brightness and the like of the moving target, and even if a certain part of the target is shielded, the tracking task can be completed as long as a part of the characteristics can be seen; it also has a good tracking effect by using in conjunction with a filter (e.g., Kalman filter).
c. Model-based tracking
The model-based tracking is to establish a model for a tracked target through manual labeling data, and then to perform feature extraction and model fitting calculation through matching the tracked target. The object tracking through the model is not easily influenced by an observation visual angle, has stronger robustness, high model matching and tracking precision, is suitable for various motion changes of a maneuvering target, and has strong anti-interference capability.
S103: and (5) detecting the personnel.
After image processing, the step needs to identify people in the video image with the determined foreground object, and comprises the following substeps:
face detection
The face detection is the first link in face recognition and is a key technology. Face detection refers to a process of determining the position, size, and posture of all faces in an image, assuming that one or more face regions exist in an input image.
The method used by the application is a new face detection algorithm which combines the face features such as skin color and the like with the iterative algorithm in order to adapt to the complexity of the background and is based on the iterative algorithm of a feature extraction algorithm (such as a Haar-like algorithm) and a cascade structure. The algorithm realizes real-time face detection, and makes the face detection technology make breakthrough progress. The algorithm firstly combines the human face characteristics, determines the approximate direction of the human face by using the human face characteristics, and then carries out verification by using the iterative algorithm of the cascade structure.
Face recognition
The application adopts a face recognition technology based on an artificial neural network. When the hybrid neural network is used for face recognition, the unsupervised neural network is used for feature extraction, and the supervised neural network is used for classification. The geometric distance between five sense organs is input into the fuzzy neural network for identification, and the effect is greatly improved compared with the common Euclidean distance-based method. Because the correlation knowledge between adjacent pixels is integrated in the convolutional neural network, invariance to image translation, rotation and local deformation is obtained to a certain extent, and a very ideal recognition result is obtained. The hybrid neural network method is better applied to the steps of face detection, face positioning and face recognition.
Voice transcription
The system adopts a real-time voice transcription technology, and is a complex calculation process for finally obtaining a text of the speaking content by utilizing a voice recognition engine to carry out feature extraction, voice decoding and language model matching on a sound signal in an audio frequency. The ASR recognition process has large calculation amount and occupies more calculation resources. Except for adopting an advanced neural network recognition algorithm, an effective measure for improving the speech recognition accuracy rate is to adopt a large amount of historical recording data for deep machine learning, the data needs to come from the field, namely real data of a national network customer service center can be used as a training set after being manually marked and checked, and the larger the training set is, the higher the recognition accuracy rate is. The identified text content is accurate, and the subsequent quality inspection and analysis algorithm can only play a role.
In the real-time voice transcription process, the key problem is how to timely process a large amount of newly generated voice every day so as to realize real-time transcription. The ASR server deployed by the system is limited, the existing computing resources are required to be fully utilized, the service volume is increased in the future, the service expansion is allowed to be carried out at any time, and the capacity increase of the system transcription capacity is completed without stopping.
S104: personnel location/statistics.
On the one hand, it is located on the basis of the person identified in step S103; and on the other hand, counting is carried out, when the statistical number of the staff in the business hall is greater than the pre-warning value set in advance, the output is carried out, and the step S108 is carried out, otherwise, the step S105 is carried out.
S108: and carrying out voice prompt alarm, starting video snapshot and simultaneously starting an emergency plan.
The emergency plan includes but is not limited to closing a gate to stop other personnel from entering; initiating personnel evacuation, etc.
For another example, when the movement track or behavior pattern of the client does not conform to the normal condition and has a large difference, an alarm and a snapshot record are automatically performed, and meanwhile, the security department is notified to manually guide the client with the abnormal behavior or take other emergency plans, so that the situation that the emotional excitement of individual client causes the personal safety of staff or other clients is prevented from occurring.
S105: and (5) tracking the personnel.
And tracking the personnel after positioning according to the movement track until the personnel leave the business hall, thereby extracting the time and the picture information of the personnel leaving the business hall.
S106: and (6) analyzing the efficiency.
Through the analysis of the real-time monitoring video, the time information of the client entering the business hall is obtained, the client is identified, positioned and tracked, and according to the analysis of the stay time in different areas, the time parameters such as the waiting time of the user for handling the business, the business handling time and the like can be obtained, so that the business handling efficiency of the business hall can be statistically analyzed.
For example, according to the number of people and the stay time in the monitoring area, the situation of people in different time periods can be determined, namely, the time corresponding to the analyzed "peak period" time period.
For another example, the total flow of people in a business hall and the number of queuing people who transact business in different time periods can be recorded to deal with the business differently. Optionally, when the number of people in line exceeds a certain value, an alarm signal is sent out to prompt the business hall to take measures such as adding a temporary business handling window and the like, so that the waiting time of clients is reduced.
For another example, by combining information such as business handling amount, business window number, business handling duration and the like of the business hall, the service bearing capacity and the overall service efficiency of the business hall are comprehensively evaluated, and reference basis is provided for optimizing business hall arrangement, business window setting and personnel configuration and improving customer experience.
S107: and (5) storing.
And storing the efficiency analysis result of the step S106 and the abnormal event information of the step S108, so as to facilitate later inquiry and retrieval.
Preferably, the method further comprises the following steps:
s109: and carrying out personalized customization based on business statistics.
Based on the historical frequency of entering and exiting of a person (client) into and from a monitoring area, the historical frequency is matched with the type and the number of business handled by the person, the business points concerned by the person are analyzed, and relevant suggestions for providing personalized and precise services for the person are provided. For example, if a client has a long time to transact a certain service, the shortest employee is scheduled to receive the service, thereby improving efficiency.
Fig. 2 is a schematic diagram illustrating the principle of the behavior supervision of the service staff in the business hall, which comprises the following steps:
s201: monitoring a video in real time;
s202: video processing;
s203: identifying personnel;
s204: personnel positioning/counting;
s205: tracking personnel;
steps S101 to S103 are the same as steps S101 to S105 in the first embodiment, and are not described again.
S206: judging whether the behavior of the staff meets the service quality inspection standard, if so, entering S207, otherwise, entering S208;
based on the conditions of arriving and leaving of the staff identified in step S103, it is determined whether the behavior of the staff in the working time meets the quality inspection standard, which specifically includes the following sub-steps:
whether the staff is late is judged according to the arrival time of the staff, and if the staff is late, the step S208 is carried out.
S207: storing and generating a service quality inspection report;
and S206, storing the behavior analysis result of the staff and the content of the video snapshot, storing event information, and generating a service quality inspection report by combining the alarm information and the video snapshot so as to investigate and collect evidence for the business hall problem complained by the customer.
S208: and performing exception handling.
The abnormal processing comprises video snapshot, the leaving time analysis is carried out according to the time nodes of leaving and returning the working position of the staff, and if the leaving time does not exceed the alarm threshold value of the leaving time, the normal leaving is judged; if the alarm threshold value of the leaving time is exceeded, alarm information prompt and video snapshot are carried out;
the method also comprises the step of starting an emergency plan aiming at the late arrival or off-duty condition of the staff, wherein the emergency plan comprises the steps of informing a manager, calling other staff and other measures so as to ensure that the staff on the station is on duty, avoiding the waiting of customers and ensuring the service quality of the business hall
S209: performing comprehensive analysis according to the service quality inspection report;
in this step, the comprehensive analysis process depends on the video analysis technology and the audio analysis technology;
the video analytics include intelligent video clip retrieval and video recall:
the video segment retrieval mainly aims at the abnormal characteristic information, time information, abnormal position information and place information input by a user, rapid screening is carried out on all index files, a final video segment address is returned, and the user can retrieve and read the retrieval result so as to carry out related quality recheck on a business hall;
the video retrieval is divided into three parts: real-time video retrieval, abnormal segment retrieval and complete video retrieval. The real-time video retrieval provides a function of viewing video pictures in each current business hall for clients; the abnormal segment retrieval aims at abnormal videos, and a user can retrieve and view the abnormal videos; the complete video retrieval can provide the function of retrieving the complete video for the user. The video retrieval comprises a video player, the detailed result of video processing is displayed, a human body or an abnormal object is marked in the image, and the marking frame moves synchronously along with the human body.
The audio analysis includes audio retrieval and audio listening:
the audio retrieval is mainly used for rapidly screening all the index storages according to the acquired time information, monitoring equipment information, abnormal information and the like, and returning the storage addresses of the audio segments, so that a user can tune, and the related quality recheck is performed on a business hall.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (8)
1. An energy efficiency improvement method based on business hall video monitoring is characterized by comprising the following steps:
A. acquiring real-time video monitoring and confirming personnel information in the video monitoring;
B. confirming the handling time of different services based on the personnel information so as to determine the handling efficiency of different employees;
C. adjusting staff allocation based on the transaction efficiency;
the personnel information includes at least identification, location and tracking of customers and employees.
2. The method according to claim 1, wherein in step a, the confirming the personnel information in the video surveillance comprises:
performing background modeling on the video monitoring image to determine a background model and a foreground image;
and carrying out steps including human shape detection, human face recognition, personnel positioning and personnel statistics in the foreground image.
3. The method of claim 1 or 2, wherein step a further comprises the step of performing speech recognition on the video surveillance.
4. The method of claim 2, wherein the step of background modeling the video surveillance image comprises: and establishing a mixed Gaussian model for the color value of each pixel point in the image, and determining a real background according to the difference of the persistence and the variability of each Gaussian distribution through sampling observation for a period of time, wherein the Gaussian distribution corresponding to the real background is used as a background model.
5. The method of claim 1, wherein step B comprises:
the method comprises the steps of identifying, positioning and tracking the client, obtaining the time of the client entering a business hall, and obtaining the waiting time and the handling time of the user for handling the business according to the time of staying in different business areas, thereby counting the business handling efficiency.
6. The method of claim 5, wherein step B further comprises: and comparing the movement track or the behavior pattern of the client with the normal movement track or the normal behavior pattern, thereby monitoring the safety behavior of the client.
7. The method according to claim 5 or 6, wherein the step B further comprises: and identifying and positioning each worker, and analyzing the leaving time according to the time node of each worker leaving and returning to the station, thereby counting the business handling efficiency of each worker.
8. The method of claim 1, wherein step C is further followed by assigning specific employees for handling specific services for different customers based on historical data of the handling efficiency;
the particular employee includes the most efficient employee to handle the particular business.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811573211.9A CN111353338B (en) | 2018-12-21 | Energy efficiency improvement method based on business hall video monitoring |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811573211.9A CN111353338B (en) | 2018-12-21 | Energy efficiency improvement method based on business hall video monitoring |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111353338A true CN111353338A (en) | 2020-06-30 |
| CN111353338B CN111353338B (en) | 2024-05-28 |
Family
ID=
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111885354A (en) * | 2020-07-15 | 2020-11-03 | 中国工商银行股份有限公司 | Service improvement discrimination method and device for bank outlets |
| CN112184497A (en) * | 2020-09-30 | 2021-01-05 | 重庆天智慧启科技有限公司 | Client visiting track tracking and passenger flow analysis system and method |
| CN112528952A (en) * | 2020-12-25 | 2021-03-19 | 合肥诚记信息科技有限公司 | Working state intelligent recognition system for electric power business hall personnel |
| CN113111847A (en) * | 2021-04-29 | 2021-07-13 | 四川隧唐科技股份有限公司 | Automatic monitoring method, device and system for process circulation |
| CN113537073A (en) * | 2021-07-19 | 2021-10-22 | 南京奥拓电子科技有限公司 | Method and system for accurately processing special events in business hall |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20060031832A (en) * | 2006-03-22 | 2006-04-13 | 뷰웨이 주식회사 | A smart visual security system based on real-time behavior analysis and situation cognizance |
| WO2006122057A2 (en) * | 2005-05-05 | 2006-11-16 | Global Management Technologies Corp. | Managing customer service staffing loads in the context of in-person transactions |
| CN101577812A (en) * | 2009-03-06 | 2009-11-11 | 北京中星微电子有限公司 | Method and system for post monitoring |
| US20120092492A1 (en) * | 2010-10-19 | 2012-04-19 | International Business Machines Corporation | Monitoring traffic flow within a customer service area to improve customer experience |
| US20150006213A1 (en) * | 2013-06-26 | 2015-01-01 | Verint Video Solutions Inc. | System and Method of Workforce Optimization |
| US9009093B1 (en) * | 2012-10-04 | 2015-04-14 | Amazon Technologies, Inc. | Deal scheduling based on user location predictions |
| CN104618685A (en) * | 2014-12-29 | 2015-05-13 | 国家电网公司 | Intelligent image analysis method for power supply business hall video monitoring |
| US20150317517A1 (en) * | 2012-11-26 | 2015-11-05 | Zte Corporation | Method, apparatus and system for acquiring headcount |
| CN106779292A (en) * | 2016-11-18 | 2017-05-31 | 浪潮（苏州）金融技术服务有限公司 | A kind of is the method and device of offering customers service based on recognition of face |
| CN106971150A (en) * | 2017-03-15 | 2017-07-21 | 国网山东省电力公司威海供电公司 | Queuing method for detecting abnormality and device that logic-based is returned |
| CN108323202A (en) * | 2017-12-27 | 2018-07-24 | 深圳前海达闼云端智能科技有限公司 | Overtime waiting prompting method and cloud system |
| WO2018132923A1 (en) * | 2017-01-20 | 2018-07-26 | Robert Johnsen | System and method for assessing customer service times |
| CN108830439A (en) * | 2018-04-09 | 2018-11-16 | 国网山东省电力公司 | Electricity business hall intelligent Service monitoring method and system based on recognition of face |
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006122057A2 (en) * | 2005-05-05 | 2006-11-16 | Global Management Technologies Corp. | Managing customer service staffing loads in the context of in-person transactions |
| KR20060031832A (en) * | 2006-03-22 | 2006-04-13 | 뷰웨이 주식회사 | A smart visual security system based on real-time behavior analysis and situation cognizance |
| CN101577812A (en) * | 2009-03-06 | 2009-11-11 | 北京中星微电子有限公司 | Method and system for post monitoring |
| US20120092492A1 (en) * | 2010-10-19 | 2012-04-19 | International Business Machines Corporation | Monitoring traffic flow within a customer service area to improve customer experience |
| US9009093B1 (en) * | 2012-10-04 | 2015-04-14 | Amazon Technologies, Inc. | Deal scheduling based on user location predictions |
| US20150317517A1 (en) * | 2012-11-26 | 2015-11-05 | Zte Corporation | Method, apparatus and system for acquiring headcount |
| US20150006213A1 (en) * | 2013-06-26 | 2015-01-01 | Verint Video Solutions Inc. | System and Method of Workforce Optimization |
| CN104618685A (en) * | 2014-12-29 | 2015-05-13 | 国家电网公司 | Intelligent image analysis method for power supply business hall video monitoring |
| CN106779292A (en) * | 2016-11-18 | 2017-05-31 | 浪潮（苏州）金融技术服务有限公司 | A kind of is the method and device of offering customers service based on recognition of face |
| WO2018132923A1 (en) * | 2017-01-20 | 2018-07-26 | Robert Johnsen | System and method for assessing customer service times |
| CN106971150A (en) * | 2017-03-15 | 2017-07-21 | 国网山东省电力公司威海供电公司 | Queuing method for detecting abnormality and device that logic-based is returned |
| CN108323202A (en) * | 2017-12-27 | 2018-07-24 | 深圳前海达闼云端智能科技有限公司 | Overtime waiting prompting method and cloud system |
| CN108830439A (en) * | 2018-04-09 | 2018-11-16 | 国网山东省电力公司 | Electricity business hall intelligent Service monitoring method and system based on recognition of face |
Non-Patent Citations (1)
| Title |
|---|
| 贾林;: "电力行业智能营业厅应用研究", 微型电脑应用, no. 11, 20 November 2012 (2012-11-20) * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111885354A (en) * | 2020-07-15 | 2020-11-03 | 中国工商银行股份有限公司 | Service improvement discrimination method and device for bank outlets |
| CN111885354B (en) * | 2020-07-15 | 2022-04-29 | 中国工商银行股份有限公司 | Service improvement discrimination method and device for bank outlets |
| CN112184497A (en) * | 2020-09-30 | 2021-01-05 | 重庆天智慧启科技有限公司 | Client visiting track tracking and passenger flow analysis system and method |
| CN112184497B (en) * | 2020-09-30 | 2023-10-03 | 重庆天智慧启科技有限公司 | Customer visit track tracking and passenger flow analysis system and method |
| CN112528952A (en) * | 2020-12-25 | 2021-03-19 | 合肥诚记信息科技有限公司 | Working state intelligent recognition system for electric power business hall personnel |
| CN112528952B (en) * | 2020-12-25 | 2022-02-11 | 合肥诚记信息科技有限公司 | Working state intelligent recognition system for electric power business hall personnel |
| CN113111847A (en) * | 2021-04-29 | 2021-07-13 | 四川隧唐科技股份有限公司 | Automatic monitoring method, device and system for process circulation |
| CN113537073A (en) * | 2021-07-19 | 2021-10-22 | 南京奥拓电子科技有限公司 | Method and system for accurately processing special events in business hall |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106997629B (en) | Access control method, apparatus and system | |
| AU2014240213B2 (en) | System and Method for object re-identification | |
| US20230289979A1 (en) | A method for video moving object detection based on relative statistical characteristics of image pixels | |
| WO2017122258A1 (en) | Congestion-state-monitoring system | |
| Guo et al. | Improved hand tracking system | |
| CN106355154B (en) | Method for detecting frequent passing of people in surveillance video | |
| Kamthe et al. | Suspicious activity recognition in video surveillance system | |
| CN110399835B (en) | Analysis method, device and system for personnel residence time | |
| CN107659754B (en) | Effective concentration method for monitoring video under condition of tree leaf disturbance | |
| CN111353343A (en) | Business hall service standard quality inspection method based on video monitoring | |
| CN109101944A (en) | A kind of real-time video monitoring algorithm identifying rubbish of jettisoninging into river | |
| WO2019114145A1 (en) | Head count detection method and device in surveillance video | |
| Le et al. | Vision-based people counting for attendance monitoring system | |
| Patil et al. | Suspicious movement detection and tracking based on color histogram | |
| CN114708555A (en) | Forest fire prevention monitoring method based on data processing and electronic equipment | |
| Yang et al. | A method of pedestrians counting based on deep learning | |
| Hirzi et al. | Literature study of face recognition using the viola-jones algorithm | |
| Lee et al. | Modeling of human walking trajectories for surveillance | |
| CN111160150A (en) | Video monitoring crowd behavior identification method based on depth residual error neural network convolution | |
| Lin et al. | An abandoned objects management system based on the gaussian mixture model | |
| Park et al. | Intensity classification background model based on the tracing scheme for deep learning based CCTV pedestrian detection | |
| CN111353338B (en) | Energy efficiency improvement method based on business hall video monitoring | |
| CN111353338A (en) | Energy efficiency improvement method based on business hall video monitoring | |
| CN114898287A (en) | Method and device for dinner plate detection early warning, electronic equipment and storage medium | |
| CN114882429A (en) | Queue counting method and system based on fusion of multiple information characteristics |
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 |