CN107341688A - The acquisition method and system of a kind of customer experience - Google Patents
The acquisition method and system of a kind of customer experience Download PDFInfo
- Publication number
- CN107341688A CN107341688A CN201710448336.8A CN201710448336A CN107341688A CN 107341688 A CN107341688 A CN 107341688A CN 201710448336 A CN201710448336 A CN 201710448336A CN 107341688 A CN107341688 A CN 107341688A
- Authority
- CN
- China
- Prior art keywords
- image
- face
- customer
- client
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06F18/2148—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the process organisation or structure, e.g. boosting cascade
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0203—Market surveys; Market polls
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/174—Facial expression recognition
Abstract
A kind of acquisition method of customer experience disclosed by the invention, comprises the following steps：S1, in real time collection client enter trade company until client leaves the overall process image of trade company；S2, using face recognition technology, pass through the customer image collected to S1 and analyze, generate customer anger index.Additionally provide the system for realizing the above method, including image capture module, the customer anger analysis module that is connected with image capture module；The advantage of the invention is that, client starts just collection customer image into trade company, when client buys product or service, the image of client is gathered in real time, analyzes customer anger, do not increased burden to client, customer experience is gathered without extra increase work, trade company is achieved that using the supervising device of itself, and client can be collected in whole process of exchange to product or the experience situation of service, it is more objective, quick, accurate, save human and material resources.
Description
Technical field
The invention belongs to enterprise's IT service management technologies field, the acquisition method and system of specially a kind of customer experience.
Background technology
With the flourishing rise of all trades and professions commercially, competition consequently also grows in intensity, and each enterprise passes through master one after another
The accessory substance of surcharge lifts competitiveness outside battalion's product, service, wherein, customer experience is that each enterprise is extremely concerned about
On one side, services client quality how is improved, customer experience satisfaction is lifted, is always the side that numerous enterprises are actively sought
To.
In order to obtain accurate customer experience in existing enterprise, obtained mostly by following several ways：1st, ecommerce
Function of Evaluation, such as the shopping website such as Taobao, Jingdone district, after client buys product or service, open Function of Evaluation, make businessman
More accurately customer experience can be obtained；2nd, enterprise it is on sale go out product or service after, by contact staff to client carry out phone
Pay a return visit, with this come obtain client for product, service satisfaction；3rd, big multi-bank industry is obtained by the way of evaluating face to face
Obtain the satisfaction of customer experience；4th, also there are some more well-known enterprises, obtained using the mode of survey for enterprise
The evaluation of popularity, the evaluation of satisfaction；But all there is objective defect for the method for above-mentioned acquisition customer experience：1st, based on visitor
Householder's dynamic formula makes evaluation, such as the Function of Evaluation of ecommerce to product or service, takes the time of client, greatly
Client does not evaluate, then obtains less than real customer experience, even phone, face to face evaluation, survey, and need client
The collection of customer experience could be completed by coordinating；2nd, the accuracy shortcoming of evaluation, the especially evaluation face to face of bank's industry, due to commenting
Valency process is too transparent so that client it is bad it is objective make evaluation, there is also this drawback for call-on back by phone, survey.
In summary, enterprise goes for customer experience more accurate and that puzzlement will not be caused to client, in existing skill
There has been no best solution in art, therefore how enterprise preferably collects customer experience, and product, service are changed with this
It is kind, or the value-added service outside lifting product, service, it is that each row of each row is all urgently to be resolved hurrily further to lift the market competitiveness
Problem.
The content of the invention
To solve drawbacks described above, the invention provides a kind of acquisition method of customer experience and system, the purpose of realization is
The satisfaction of customer experience can be obtained according to the mood of client, client gathers in real time while product or service is bought, no
Client can be caused to perplex, enterprise being individually acquired also because without having saved manpower, goods and materials cost.
To achieve these goals, the present invention provides following technical scheme：A kind of customer experience provided by the invention is adopted
Diversity method, comprise the following steps：S1, in real time collection client enter trade company until the overall process image that client leaves；
S2, using face recognition technology, pass through the customer image collected to S1 and analyze, generate customer anger index.
Start just collection customer image into trade company in client, when client buys product or service, gather client in real time
Image, analyze customer anger, do not increased burden to client, trade company achieves that using the supervising device of itself, is merchandising
During can collect client to product or the experience situation of service, it is quick, accurate, save human and material resources.
Further, the S2 includes：
S21, establish face mood expression data storehouse；
S22, face features are detected using face recognition technology, accurately mark position and the size of face, obtain visitor
Family face-image, is stored after pretreatment, then carries out facial image feature extraction, the facial image feature and face that will be extracted
Facial expression image in mood expression data storehouse compares, and generates customer anger index.
The face mood expression data storehouse preferably uses The Extended Cohn-Kanade Dataset, abbreviation CK
+ database, this database are to extend to come on the basis of Cohn-Kanade Dataset, are published on 2010.This number
According to storehouse greatly more are wanted compared with JAFFE.And can also Free Acquisition, label the and Action Units' comprising expression
label.This database includes 123 subjects, and 593 image sequence, each image sequence are most
Latter Frame has action units label, and in this 593 image sequence, there are 327
Sequence has emotion label.A total of 7 groups of facial expression images of the database, wherein file 1 represent angry facial expression, 2 tables
Show and despise expression, 3 square one's shoulders expression, and 4 represent frightened expressions, and 5 represent smile expressions, and 6 represent sad expressions, and 7 express surprise
Expression, by the client's face-image collected after face recognition chip comes out the feature extraction that can represent human face expression,
Facial expression image directly with lane database contrasts, and occupancy resource is few, calculating speed is fast, and the degree of accuracy is high.
Further, client's facial expression image is obtained, by client face by facial image feature extraction in the S22
Portion's facial expression image compares one by one with the facial expression image in face mood expression data storehouse, choose in face mood expression data storehouse with
Client's facial expression similarity highest facial expression image, the as sentiment indicator corresponding to the facial expression image, client's current emotional
Index.Comparison method is simple, and the degree of accuracy is high, speed is fast.
Further, by the facial expression image in client's facial expression image and face mood expression data storehouse in the S22
When comparing one by one, using Extraction of Geometrical Features method, respectively by client's facial expression image, face mood expression data storehouse
Mouth, eyebrow, nose in facial expression image, eyes these human face expressions notable feature shape and change in location positioned and
Measurement, determines mouth in respective image, eyebrow, nose, eye shape, size, distance and mutual ratio, finally by client's face
Mouth, eyebrow, nose, eye shape, size, distance and mutual ratio and the expression in face mood expression data storehouse in facial expression image
Mouth, eyebrow, nose, eye shape, size, distance and mutual ratio in image are compared.In specific be compared, choosing
Take some emphasis to represent the parameter at the face position of expression to be compared, compare that speed is fast, objective contrasts the comparison for being
Result precision is high.
Further, the S21 assigns numerical value in face mood expression data storehouse to every kind of sentiment indicator；
After customer anger index is obtained in the S22, establish and trade company entered using client until leaving trade company's time as transverse axis,
The numerical value that customer anger index assigns is the functional relation of the longitudinal axis, and generates visualization display.Obtained functional relation is more straight
That sees recognizes that client enters trade company until leaving the change of trade company's sentiment indicator, and trade company can enter one according to the time of interaction
The change of step analysis customer anger index, such as can be entering the sentiment indicator statistics of trade company's time, collecting, in purchase product or clothes
The sentiment indicator being engaged in this period is counted, collected, and is left the sentiment indicator statistics of trade company's time, is collected, more accurate with this
Obtain when trade company obtains product or service and more accurately experience satisfaction.
Further, when client's face-image is obtained in the S22, in addition to the customer image obtained from S1 is divided in real time
Frame of video is analysed, human face image sequence, the time for gathering image and collection client's face-image device are extracted from frame of video
IP address information.
Further, the time according to collection image is also included in the S22, gathers the IP address letter of customer image device
Breath carries out the statistical analysis of time section, band of position customer anger index.
Further, also include creating trade company's input area in the S21, trade company can input the visitor obtained with customer communication
Family sentiment indicator, in the upload of face mood expression data storehouse addition trade company, the client face image chosen and customer anger index, enter
Row deep learning.In order to make the degree of accuracy of collection customer experience higher, the present invention can also be by face mood expression data storehouse
Deep learning, so as to obtain database more horn of plenty, trade company's using flexible is high.
Present invention also offers the system for realizing the above method, including image capture module, it is connected with image capture module
Customer anger analysis module；
Image capture module is used to gather client in real time into trade company until client leaves the overall process image of trade company；
Customer anger analysis module is used to, by analyzing the client's face-image collected, generate customer anger index.
Further, the customer anger analysis module includes face mood expression data storehouse, face recognition chip, comparison
Module, generation customer anger Index module；
Face recognition chip includes the face detection module, acquisition client's face-image module, facial image being connected with each other
Pretreatment module, facial image characteristic extracting module；
Face mood expression data storehouse is used to establish face mood expression data storehouse；
Face recognition chip is used to detect face features, accurately marks position and the size of face, obtains client face
Portion's image, is stored after pretreatment, then carries out facial image feature extraction；
Comparing module is used to do the facial image feature extracted and the facial expression image in face mood expression data storehouse
Compare；
Generation customer anger Index module is used for the result for receiving comparing module, and result is carried out into visualization and shown.
Further, face mood expression data storehouse includes setting sentiment indicator module, sets sentiment indicator module by people
Every kind of sentiment indicator assigns numerical value in face mood expression data storehouse；
Generating customer anger Index module includes analysis customer anger trend module, and analysis customer anger trend module is used for
After obtaining customer anger index, establish and trade company is entered using client until leaving trade company's time as transverse axis, customer anger index assigns
Numerical value be the longitudinal axis functional relation, and generate visualization display.
Further, generating customer anger Index module includes analysis video frame module in real time, analyzes frame of video mould in real time
Block is used to from client's face-image of acquisition extract human face image sequence from frame of video, gathers the time of image and adopt
Collect the IP address information of client's face-image device.
Further, the customer anger that customer anger Index module includes with analysis video frame module is connected in real time is generated to refer to
Statistical analysis module is marked, customer anger indicator-specific statistics analysis module is used for according to the face obtained from real-time analysis video frame module
Image sequence, the time for gathering image, the IP address information progress time section for gathering client's face-image device, the band of position
The statistical analysis of customer anger index.
Further, face mood expression data storehouse includes trade company's input module, inputs for trade company and is obtained with customer communication
The customer anger index arrived, the upload of addition trade company, the client face image chosen and corresponding sentiment indicator, makes face mood expression
Database carries out deep learning.
The present invention uses above-mentioned technical proposal, including following beneficial effect：Client starts just collection client figure into trade company
Picture, when client buys product or service, the image of client is gathered in real time, customer anger is analyzed, is not increased burden to client,
Customer experience is gathered without additionally increasing work, trade company achieves that using the supervising device of itself, is merchandised entirely
Client can be collected in journey to product or the experience situation of service, more objective, quick, accurate, saving human and material resources.
Brief description of the drawings
Fig. 1 is customer anger index trend curve figure in example IV.
Embodiment
The present invention is described in further detail below by specific embodiment and with reference to accompanying drawing.
Embodiment one：A kind of acquisition method of customer experience provided by the invention, comprises the following steps：
S1, in real time collection client enter trade company until client leaves the overall process image of trade company；
S2, using face recognition technology, pass through the customer image collected to S1 and analyze, generate customer anger index.
The system for realizing the above method, including image capture module, be connected with image capture module customer anger analysis
Module；
Image capture module is used to gather client in real time into trade company until client leaves the overall process image of trade company；
Customer anger analysis module is used to, by analyzing the customer image collected, generate customer anger index.
Can use video camera, video recorder these have camera function device collection client's face-image, the step S1,
S2 has no strict time sequencing, can carry out simultaneously, when gathering face-image, just often uses face recognition technology simultaneously
, face face-image is analyzed in collection, embodies flat, excited, angry sentiment indicator.
This method that the present invention designs, integrate with client trading, collection customer experience, without other extra steps
It is rapid that customer experience is acquired, save human and material resources.
Embodiment two：A kind of acquisition method of customer experience provided by the invention, comprises the following steps：
S1, in real time collection client enter trade company until client leaves the overall process image of trade company；
S2, using face recognition technology, pass through the customer image collected to S1 and analyze, generate customer anger index.
Further, the S2 includes：
S21, establish face mood expression data storehouse；
S22, face features are detected using face recognition technology, accurately mark position and the size of face, obtain visitor
Family face-image, is stored after pretreatment, then carries out facial image feature extraction, the facial image feature and face that will be extracted
Facial expression image in mood expression data storehouse compares, and generates customer anger index.
The system for realizing the above method, including image capture module, be connected with image capture module customer anger analysis
Module；
Image capture module is used to gather client in real time into trade company until client leaves the overall process image of trade company；
Customer anger analysis module is used to, by analyzing the customer image collected, generate customer anger index.
Further, the customer anger analysis module includes face mood expression data storehouse, face recognition chip, comparison
Module, generation customer anger Index module；
Face recognition chip includes the face detection module, acquisition client's face-image module, facial image being connected with each other
Pretreatment module, facial image characteristic extracting module；
Face mood expression data storehouse is used to establish face mood expression data storehouse；
Face recognition chip is used to detect face features, accurately marks position and the size of face, obtains client face
Portion's image, is stored after pretreatment, then carries out facial image feature extraction；
Comparing module is used to do the facial image feature extracted and the facial expression image in face mood expression data storehouse
Compare；
Generation customer anger Index module is used for the result for receiving comparing module, and result is carried out into visualization and shown.
The Extended Cohn-Kanade Dataset may be selected in the face mood expression data storehouse established, referred to as
CK+ databases, wherein a total of 7 groups of facial expression images of database, file 1 represent angry facial expression, and 2 represent to despise expression, and 3 represent to detest
Dislike expression, 4 represent frightened expressions, and 5 represent smile expressions, and 6 represent sad expressions, and 7 express surprise expression, face recognition technology one
As and S1 collection client's face-images carry out simultaneously, using face recognition technology, the facial image of S1 collections is analyzed, generation is objective
The idiographic flow of family sentiment indicator is：Face features are first detected, distinguish background, face, for camera precise acquisition
To face, different facial images can be transferred through pick-up lens and collect, such as still image, dynamic image, different positions
Put, different expressions etc. can be gathered well.When user is in the coverage of collecting device, collecting device
It can automatically search for and shoot the facial image of user；
Then accurate calibration goes out position and the size of face in the picture, and the pattern feature included in facial image is very rich
Richness, such as histogram feature, color characteristic, template characteristic, architectural feature and Haar features.Face detection module is exactly this its
In useful information pick out, and realize Face datection using these features.
The method for detecting human face of main flow uses Adaboost learning algorithms based on features above, and Adaboost algorithm is a kind of
For the method classified, it is combined some weaker sorting techniques, is combined into new very strong sorting technique.
Rectangular characteristic (the weak typing of face can most be represented by picking out some using Adaboost algorithm during Face datection
Device), Weak Classifier is configured to a strong classifier in the way of Nearest Neighbor with Weighted Voting, then obtained some strong classifiers will be trained
The cascade filtering of a cascade structure is composed in series, effectively improves the detection speed of grader；
Obtain client's face-image module and get client's face-image that S1 is collected, facial image pretreatment module pair
It is to be based on Face datection result in the image preprocessing of face, image is handled and finally serves the mistake of feature extraction
Journey.The original image that system obtains tends not to directly use due to being limited and random disturbances by various conditions, it is necessary to
The early stage of image procossing carries out the image preprocessings such as gray correction, noise filtering to it.For facial image, its is pre-
The processing procedure mainly light compensation including facial image, greyscale transformation, histogram equalization, normalization, geometric correction, filtering
And sharpen etc.；
Facial image characteristic extracting module：Feature workable for face identification system is generally divided into visual signature, pixel system
Count feature, facial image conversion coefficient feature, facial image algebraic characteristic etc..Face characteristic is extracted aiming at some of face
What feature was carried out.Face characteristic extracts, and also referred to as face characterizes, and it is the process that feature modeling is carried out to face.Face characteristic carries
The method taken, which is summed up, is divided into two major classes：One kind is Knowledge based engineering characterizing method；Another be based on algebraic characteristic or
The characterizing method of statistical learning.
Finally, comparing module does the facial image feature extracted and the facial expression image in face mood expression data storehouse
Compare, generate customer anger index, this step mainly apply recognition of face in matching step, such as with surprised table in database
Feelings group is matched, and client's current emotional is obtained according to the similarity of matching, typically can all be set a threshold value in matching, be exceeded
Threshold value or highest beyond threshold value match the most, for example, be also classified into surprised expression group it is general it is surprised, moderate is surprised,
It is highly surprised, and three kinds of degree are matched, moderate is surprised beyond matching threshold, then it is moderate to be determined as client's current emotional
It is surprised, through generating the generation visualization display of customer anger Index module.
Embodiment three：A kind of acquisition method of customer experience provided by the invention, comprises the following steps：
S1, collection customer enters trade company until client leaves the overall process image of trade company in real time；
S2, using face recognition technology, pass through the customer image collected to S1 and analyze, generate customer anger index.
Further, the S2 includes：
S21, establish face mood expression data storehouse；
S22, face features are detected using face recognition technology, accurately mark position and the size of face, obtain visitor
Family face-image, is stored after pretreatment, then carries out facial image feature extraction, the facial image feature and face that will be extracted
Facial expression image in mood expression data storehouse compares, and generates customer anger index；Carried in the S22 by facial image feature
Take, obtain client's facial expression image, by the facial expression image in client's facial expression image and face mood expression data storehouse by
One compares, choose in face mood expression data storehouse with client's facial expression similarity highest facial expression image, the facial expression image
Corresponding sentiment indicator, as client's current emotional index.
Specifically, in the S22 by the facial expression image in client's facial expression image and face mood expression data storehouse by
One when comparing, using Extraction of Geometrical Features method, respectively by client's facial expression image, the table in face mood expression data storehouse
Mouth, eyebrow, nose in feelings image, eyes these human face expressions notable feature shape and change in location positioned and surveyed
Amount, determines mouth in respective image, eyebrow, nose, eye shape, size, distance and mutual ratio, finally by client's face table
Mouth, eyebrow, nose, eye shape, size, distance and mutual ratio and the expression figure in face mood expression data storehouse in feelings image
Mouth, eyebrow, nose, eye shape, size, distance and mutual ratio as in are compared.Compared using above-mentioned parameter
It is right, it can further improve the accuracy rate of comparing result.
Completing the system of the above method includes image capture module, the customer anger being connected with image capture module analysis mould
Block；
Image capture module is used to gather client in real time into trade company until client leaves the overall process image of trade company；
Customer anger analysis module is used to, by analyzing the customer image collected, generate customer anger index.
The customer anger analysis module includes face mood expression data storehouse, face recognition chip, comparing module, generation
Customer anger Index module；
Face recognition chip includes the face detection module, acquisition client's face-image module, facial image being connected with each other
Pretreatment module, facial image characteristic extracting module；
Face mood expression data storehouse is used to establish face mood expression data storehouse；
Face recognition chip is used to detect face features, accurately marks position and the size of face, obtains client face
Portion's image, is stored after pretreatment, then carries out facial image feature extraction；
Comparing module is used to do the facial image feature extracted and the facial expression image in face mood expression data storehouse
Compare；
Generation customer anger Index module is used for the result for receiving comparing module, and result is carried out into visualization and shown.
Example IV：A kind of acquisition method of customer experience provided by the invention, comprises the following steps：
S1, in real time collection client enter trade company until client leaves the overall process image of trade company；
S2, using face recognition technology, pass through the customer image collected to S1 and analyze, generate customer anger index.
Further, the S2 includes：
S21, establish face mood expression data storehouse；
S22, face features are detected using face recognition technology, accurately mark position and the size of face, obtain visitor
Family face-image, is stored after pretreatment, then carries out facial image feature extraction, the facial image feature and face that will be extracted
Facial expression image in mood expression data storehouse compares, and generates customer anger index；Carried in the S22 by facial image feature
Take, obtain client's facial expression image, by the facial expression image in client's facial expression image and face mood expression data storehouse by
One compares, choose in face mood expression data storehouse with client's facial expression similarity highest facial expression image, the facial expression image
Corresponding sentiment indicator, as client's current emotional index.
Specifically, in the S22 by the facial expression image in client's facial expression image and face mood expression data storehouse by
One when comparing, using Extraction of Geometrical Features method, respectively by client's facial expression image, the table in face mood expression data storehouse
Mouth, eyebrow, nose in feelings image, eyes these human face expressions notable feature shape and change in location positioned and surveyed
Amount, determines mouth in respective image, eyebrow, nose, eye shape, size, distance and mutual ratio, finally by client's face table
Mouth, eyebrow, nose, eye shape, size, distance and mutual ratio and the expression figure in face mood expression data storehouse in feelings image
Mouth, eyebrow, nose, eye shape, size, distance and mutual ratio as in are compared.
Further, the S21 assigns numerical value in face mood expression data storehouse to every kind of sentiment indicator；
After customer anger index is obtained in the S22, establish and trade company entered using client until leaving trade company's time as transverse axis,
The numerical value that customer anger index assigns is the functional relation of the longitudinal axis, and generates visualization display.
For example, the face mood expression data storehouse selection The Extended Cohn-Kanade Dataset established,
Abbreviation CK+ databases, wherein a total of 7 groups of facial expression images of database, file 1 represent angry facial expression, and 2 represent to despise expression, 3 tables
Show detest expression, 4 represent frightened expressions, and 5 represent smile expressions, and 6 represent sad expressions, and 7 express surprise expression, respectively to every group
Mood assigns numerical value, as shown in the table：
To gather time of client's face-image as transverse axis, the numerical value of above-mentioned imparting is the longitudinal axis, obtains customer anger index
Trend curve, as shown in figure 1, the image collected with reference to S1, so that it may understand client from the mood for entering trade company, seeing commodity
Mood, more can accurately actively distinguish that client is that product or service are unsatisfied with during transaction, is still had just enter into
Trade company itself is in low spirits, or is unsatisfied with during being talked with trade company seller, can also analyze client to that product
Favor and dislike.
Completing the system of the above method includes image capture module, the customer anger being connected with image capture module analysis mould
Block；
Image capture module is used to gather client in real time into trade company until client leaves the overall process image of trade company；
Customer anger analysis module is used to, by analyzing the customer image collected, generate customer anger index.
The customer anger analysis module includes face mood expression data storehouse, face recognition chip, comparing module, generation
Customer anger Index module；
Face recognition chip includes the face detection module, acquisition client's face-image module, facial image being connected with each other
Pretreatment module, facial image characteristic extracting module；
Face mood expression data storehouse is used to establish face mood expression data storehouse；
Face recognition chip is used to detect face features, accurately marks position and the size of face, obtains client face
Portion's image, is stored after pretreatment, then carries out facial image feature extraction；
Comparing module is used to do the facial image feature extracted and the facial expression image in face mood expression data storehouse
Compare；
Generation customer anger Index module is used for the result for receiving comparing module, and result is carried out into visualization and shown.
Face mood expression data storehouse includes setting sentiment indicator module, sets sentiment indicator module by face mood expression
Every kind of sentiment indicator assigns numerical value in database；
Generating customer anger Index module includes analysis customer anger trend module, and analysis customer anger trend module is used for
After obtaining customer anger index, establish and trade company is entered using client until leaving trade company's time as transverse axis, customer anger index assigns
Numerical value be the longitudinal axis functional relation, and generate visualization display.
Embodiment five：A kind of acquisition method of customer experience provided by the invention, comprises the following steps：
S1, in real time collection client enter trade company until client leaves the overall process image of trade company；
S2, using face recognition technology, pass through the customer image collected to S1 and analyze, generate customer anger index.
Further, the S2 includes：
S21, establish face mood expression data storehouse；
S22, face features are detected using face recognition technology, accurately mark position and the size of face, obtain visitor
Family face-image, is stored after pretreatment, then carries out facial image feature extraction, the facial image feature and face that will be extracted
Facial expression image in mood expression data storehouse compares, and generates customer anger index；Carried in the S22 by facial image feature
Take, obtain client's facial expression image, by the facial expression image in client's facial expression image and face mood expression data storehouse by
One compares, choose in face mood expression data storehouse with client's facial expression similarity highest facial expression image, the facial expression image
Corresponding sentiment indicator, as client's current emotional index.
Specifically, in the S22 by the facial expression image in client's facial expression image and face mood expression data storehouse by
One when comparing, using Extraction of Geometrical Features method, respectively by client's facial expression image, the table in face mood expression data storehouse
Mouth, eyebrow, nose in feelings image, eyes these human face expressions notable feature shape and change in location positioned and surveyed
Amount, determines mouth in respective image, eyebrow, nose, eye shape, size, distance and mutual ratio, finally by client's face table
Mouth, eyebrow, nose, eye shape, size, distance and mutual ratio and the expression figure in face mood expression data storehouse in feelings image
Mouth, eyebrow, nose, eye shape, size, distance and mutual ratio as in are compared.
Further, the S21 assigns numerical value in face mood expression data storehouse to every kind of sentiment indicator；
After customer anger index is obtained in the S22, establish and trade company entered using client until leaving trade company's time as transverse axis,
The numerical value that customer anger index assigns is the functional relation of the longitudinal axis, and generates visualization display.
Further, when client's face-image is obtained in the S22, in addition to the client's face-image obtained from S1 is real
When analyze frame of video, human face image sequence, the time for gathering image and collection customer image device are extracted from frame of video
IP address information.
When also including the time according to collection image in the S22, gathering the IP address information progress of customer image device
Between section, band of position customer anger index statistical analysis.
When general camera sends video flowing, which camera is sent during in order to distinguish, and can all assign an IP address,
Physical spatial location is corresponded to according to IP address, with reference to acquisition time, using clear data, you can statistical separates out and for example has just enter into business
This period of family, the customer anger index of this period of shopping goods, make analysis more easy.
Further, also include creating trade company's input area in the S21, trade company can input the visitor obtained with customer communication
Family sentiment indicator, in the upload of face mood expression data storehouse addition trade company, the client face image chosen and customer anger index, enter
Row deep learning.Trade company can use this step to choose image when being linked up with trade company personnel from S1, then pass through trade company personnel
The sentiment indicator description linked up and recognized is crossed, increase facial expression image, the sentiment indicator of database, database depth is made with this
Practise, further obtain more accurate customer anger index.
Completing the system of the above method includes image capture module, the customer anger being connected with image capture module analysis mould
Block；
Image capture module is used to gather client in real time into trade company until client leaves the image of the overall process of trade company；
Customer anger analysis module is used to, by analyzing the customer image collected, generate customer anger index.
The customer anger analysis module includes face mood expression data storehouse, face recognition chip, comparing module, generation
Customer anger Index module；
Face recognition chip includes the face detection module, acquisition client's face-image module, facial image being connected with each other
Pretreatment module, facial image characteristic extracting module；
Face mood expression data storehouse is used to establish face mood expression data storehouse；
Face recognition chip is used to detect face features, accurately marks position and the size of face, obtains client face
Portion's image, is stored after pretreatment, then carries out facial image feature extraction；
Comparing module is used to do the facial image feature extracted and the facial expression image in face mood expression data storehouse
Compare；
Generation customer anger Index module is used for the result for receiving comparing module, and result is carried out into visualization and shown.
Face mood expression data storehouse includes setting sentiment indicator module, sets sentiment indicator module by face mood expression
Every kind of sentiment indicator assigns numerical value in database；
Generating customer anger Index module includes analysis customer anger trend module, and analysis customer anger trend module is used for
After obtaining customer anger index, establish and trade company is entered using client until leaving trade company's time as transverse axis, customer anger index assigns
Numerical value be the longitudinal axis functional relation, and generate visualization display.
Generating customer anger Index module includes analysis video frame module in real time, analyzes video frame module in real time and is used for from obtaining
Human face image sequence, the time for gathering image and collection customer image are extracted in the client's face-image taken from frame of video
The IP address information of device.
Generation customer anger Index module includes the customer anger indicator-specific statistics point being connected with analysis video frame module in real time
Module is analysed, customer anger indicator-specific statistics analysis module is used for according to the facial image sequence obtained from real-time analysis video frame module
Row, the time of collection image, the IP address information of collection client's face-image device carry out time section, band of position client's feelings
The statistical analysis of thread index.
Face mood expression data storehouse includes trade company's input module, the client's feelings obtained for trade company's input with customer communication
Thread index, the upload of addition trade company, the client face image chosen and corresponding sentiment indicator, carries out face mood expression data storehouse
Deep learning.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention, for the skill of this area
For art personnel, the present invention can have various modifications and variations.Within the spirit and principles of the invention, that is made any repaiies
Change, equivalent substitution, improvement etc., should be included in the scope of the protection.
Claims (10)
1. a kind of acquisition method of customer experience, it is characterised in that this method comprises the following steps：
S1, in real time collection client enter trade company until the overall process image that client leaves；
S2, using face recognition technology, pass through the customer image collected to S1 and analyze, generate customer anger index.
2. the acquisition method of customer experience according to claim 1, it is characterised in that the S2 includes：
S21, establish face mood expression data storehouse；
S22, face features are detected using face recognition technology, accurately mark position and the size of face, obtain client face
Portion's image, is stored after pretreatment, then carries out facial image feature extraction, by the facial image feature extracted and face mood
Facial expression image in expression data storehouse compares, and generates customer anger index.
3. the acquisition method of customer experience according to claim 2, it is characterised in that special by facial image in the S22
Sign extraction, obtains client's facial expression image, by the expression figure in client's facial expression image and face mood expression data storehouse
As comparing one by one, choose in face mood expression data storehouse with client's facial expression similarity highest facial expression image, the expression
Sentiment indicator corresponding to image, as client's current emotional index.
4. the acquisition method of customer experience according to claim 3, it is characterised in that by client's facial expression in the S22
When image compares one by one with the facial expression image in face mood expression data storehouse, using Extraction of Geometrical Features method, respectively by visitor
Mouth, eyebrow, nose in the facial expression image of family, in the facial expression image in face mood expression data storehouse, eyes these face tables
The shape and change in location of the notable feature of feelings are positioned and measured, and determine mouth, eyebrow, nose, eyes in respective image
Shape, size, distance and mutual ratio, finally by mouth in client's facial expression image, eyebrow, nose, eye shape, size,
Mouth, eyebrow, nose, eye shape, size in distance and the mutually facial expression image in ratio and face mood expression data storehouse, away from
From and mutually ratio be compared.
5. the acquisition method of customer experience according to claim 2, it is characterised in that the S21 is in face mood expression number
According in storehouse, numerical value is assigned to every kind of sentiment indicator；
After customer anger index is obtained in the S22, establish and trade company is entered using client until leaving trade company's time as transverse axis, client
The numerical value that sentiment indicator assigns is the functional relation of the longitudinal axis, and generates visualization display.
6. the acquisition method of customer experience according to claim 2, it is characterised in that client's face figure is obtained in the S22
During picture, in addition to the customer image obtained from S1 analyzed into frame of video in real time, human face image sequence, collection are extracted from frame of video
The time of image and the IP address information for gathering customer image device.
7. the acquisition method of customer experience according to claim 6, it is characterised in that also include in the S22 according to collection
The time of image, the IP address information progress time section for gathering customer image device, the system of band of position customer anger index
Meter analysis.
8. the acquisition method of customer experience according to claim 2, it is characterised in that also include creating trade company in the S21
Input area, trade company can input the customer anger index obtained with customer communication, and trade company is added in face mood expression data storehouse
The client face image and customer anger index upload, chosen, carry out deep learning.
9. a kind of acquisition system of customer experience, it is characterised in that the system includes image capture module and image capture module
The customer anger analysis module of connection；
Image capture module is used to gather client in real time into trade company until client leaves the overall process image of trade company；
Customer anger analysis module is used to, by analyzing the customer image collected, generate customer anger index.
10. the acquisition system of customer experience according to claim 9, it is characterised in that the customer anger analysis module bag
Include face mood expression data storehouse, face recognition chip, comparing module, generation customer anger Index module；
Face recognition chip includes the face detection module of interconnection, acquisition client's face-image module, facial image and located in advance
Manage module, facial image characteristic extracting module；
Face mood expression data storehouse is used to establish face mood expression data storehouse；
Face recognition chip is used to detect face features, accurately marks position and the size of face, obtains client's face figure
Picture, store after pretreatment, then carry out facial image feature extraction；
Comparing module is used to compare the facial image feature extracted and the facial expression image in face mood expression data storehouse；
Generation customer anger Index module is used for the result for receiving comparing module, and result is carried out into visualization and shown；
Face mood expression data storehouse includes setting sentiment indicator module, sets sentiment indicator module by face mood expression data
Every kind of sentiment indicator assigns numerical value in storehouse；
Generating customer anger Index module includes analysis customer anger trend module, and analysis customer anger trend module is used to obtain
After customer anger index, establish and trade company is entered using client until leaving trade company's time as transverse axis, the number that customer anger index assigns
It is worth the functional relation for the longitudinal axis, and generates visualization display；
Generating customer anger Index module includes analysis video frame module in real time, analyzes video frame module in real time and is used for from acquisition
Human face image sequence, the time for gathering image and collection customer image device are extracted in client's face-image from frame of video
IP address information；
The customer anger indicator-specific statistics that generation customer anger Index module includes with analysis video frame module is connected in real time analyzes mould
Block, customer anger indicator-specific statistics analysis module are used for according to the human face image sequence obtained from real-time analysis video frame module, adopted
The time for collecting image, the IP address information for gathering client's face-image device carry out time section, band of position customer anger refers to
Target statistical analysis；
Face mood expression data storehouse includes trade company's input module, and the customer anger obtained for trade company's input with customer communication refers to
Mark, the upload of addition trade company, the client face image chosen and corresponding sentiment indicator, makes face mood expression data storehouse carry out depth
Study.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710448336.8A CN107341688A (en) | 2017-06-14 | 2017-06-14 | The acquisition method and system of a kind of customer experience |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710448336.8A CN107341688A (en) | 2017-06-14 | 2017-06-14 | The acquisition method and system of a kind of customer experience |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107341688A true CN107341688A (en) | 2017-11-10 |
Family
ID=60221061
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710448336.8A Pending CN107341688A (en) | 2017-06-14 | 2017-06-14 | The acquisition method and system of a kind of customer experience |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107341688A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108509941A (en) * | 2018-04-20 | 2018-09-07 | 北京京东金融科技控股有限公司 | Emotional information generation method and device |
| CN108564007A (en) * | 2018-03-27 | 2018-09-21 | 深圳市智能机器人研究院 | A kind of Emotion identification method and apparatus based on Expression Recognition |
| CN108860820A (en) * | 2018-06-28 | 2018-11-23 | 重庆瑞芝电子商务有限公司 | processing system for commodity packaging |
| CN109784977A (en) * | 2018-12-18 | 2019-05-21 | 深圳壹账通智能科技有限公司 | Service methods of marking, device, computer equipment and storage medium |
| CN109919102A (en) * | 2019-03-11 | 2019-06-21 | 重庆科技学院 | A kind of self-closing disease based on Expression Recognition embraces body and tests evaluation method and system |
| CN109919099A (en) * | 2019-03-11 | 2019-06-21 | 重庆科技学院 | A kind of user experience evaluation method and system based on Expression Recognition |
| CN109920514A (en) * | 2019-03-11 | 2019-06-21 | 重庆科技学院 | A kind of self-closing disease based on Kalman filtering neural network embraces body and tests evaluation method and system |
| CN109919100A (en) * | 2019-03-11 | 2019-06-21 | 重庆科技学院 | A kind of user experience evaluation method and system based on cell phone client Yu cloud service technology |
| CN110378736A (en) * | 2019-07-23 | 2019-10-25 | 中国科学院东北地理与农业生态研究所 | The method that tourist experiences satisfaction to natural resources is evaluated by facial expression recognition |
| CN111802320A (en) * | 2019-04-12 | 2020-10-23 | 重庆医科大学 | Tail clamping device for construction of depression animal model, expression image acquisition system, model construction method and application |
| CN112036328A (en) * | 2020-09-02 | 2020-12-04 | 中国银行股份有限公司 | Bank customer satisfaction calculation method and device |
| CN112183417A (en) * | 2020-09-30 | 2021-01-05 | 重庆天智慧启科技有限公司 | Business consultant service capability evaluation system and method |
| CN112200051A (en) * | 2020-09-30 | 2021-01-08 | 重庆天智慧启科技有限公司 | Case field inspection system and method |
| CN112581230A (en) * | 2020-12-24 | 2021-03-30 | 安徽航天信息科技有限公司 | Commodity recommendation method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101998052A (en) * | 2009-08-07 | 2011-03-30 | 奥林巴斯映像株式会社 | Photographing apparatus |
| CN102831447A (en) * | 2012-08-30 | 2012-12-19 | 北京理工大学 | Method for identifying multi-class facial expressions at high precision |
| CN104699769A (en) * | 2015-02-28 | 2015-06-10 | 北京京东尚科信息技术有限公司 | Interacting method based on facial expression recognition and equipment executing method |
| US20170046561A1 (en) * | 2013-08-19 | 2017-02-16 | Kodak Alaris Inc. | Imaging workflow using facial and non-facial features |
| CN106485227A (en) * | 2016-10-14 | 2017-03-08 | 深圳市唯特视科技有限公司 | A kind of Evaluation of Customer Satisfaction Degree method that is expressed one's feelings based on video face |
-
2017
- 2017-06-14 CN CN201710448336.8A patent/CN107341688A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101998052A (en) * | 2009-08-07 | 2011-03-30 | 奥林巴斯映像株式会社 | Photographing apparatus |
| CN102831447A (en) * | 2012-08-30 | 2012-12-19 | 北京理工大学 | Method for identifying multi-class facial expressions at high precision |
| US20170046561A1 (en) * | 2013-08-19 | 2017-02-16 | Kodak Alaris Inc. | Imaging workflow using facial and non-facial features |
| CN104699769A (en) * | 2015-02-28 | 2015-06-10 | 北京京东尚科信息技术有限公司 | Interacting method based on facial expression recognition and equipment executing method |
| CN106485227A (en) * | 2016-10-14 | 2017-03-08 | 深圳市唯特视科技有限公司 | A kind of Evaluation of Customer Satisfaction Degree method that is expressed one's feelings based on video face |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108564007B (en) * | 2018-03-27 | 2021-10-22 | 深圳市智能机器人研究院 | Emotion recognition method and device based on expression recognition |
| CN108564007A (en) * | 2018-03-27 | 2018-09-21 | 深圳市智能机器人研究院 | A kind of Emotion identification method and apparatus based on Expression Recognition |
| CN108509941A (en) * | 2018-04-20 | 2018-09-07 | 北京京东金融科技控股有限公司 | Emotional information generation method and device |
| CN108860820A (en) * | 2018-06-28 | 2018-11-23 | 重庆瑞芝电子商务有限公司 | processing system for commodity packaging |
| CN109784977A (en) * | 2018-12-18 | 2019-05-21 | 深圳壹账通智能科技有限公司 | Service methods of marking, device, computer equipment and storage medium |
| CN109919102A (en) * | 2019-03-11 | 2019-06-21 | 重庆科技学院 | A kind of self-closing disease based on Expression Recognition embraces body and tests evaluation method and system |
| CN109919099A (en) * | 2019-03-11 | 2019-06-21 | 重庆科技学院 | A kind of user experience evaluation method and system based on Expression Recognition |
| CN109920514A (en) * | 2019-03-11 | 2019-06-21 | 重庆科技学院 | A kind of self-closing disease based on Kalman filtering neural network embraces body and tests evaluation method and system |
| CN109919100A (en) * | 2019-03-11 | 2019-06-21 | 重庆科技学院 | A kind of user experience evaluation method and system based on cell phone client Yu cloud service technology |
| CN111802320A (en) * | 2019-04-12 | 2020-10-23 | 重庆医科大学 | Tail clamping device for construction of depression animal model, expression image acquisition system, model construction method and application |
| CN110378736A (en) * | 2019-07-23 | 2019-10-25 | 中国科学院东北地理与农业生态研究所 | The method that tourist experiences satisfaction to natural resources is evaluated by facial expression recognition |
| CN110378736B (en) * | 2019-07-23 | 2023-01-03 | 中国科学院东北地理与农业生态研究所 | Method for evaluating experience satisfaction degree of tourists on natural resources through facial expression recognition |
| CN112036328A (en) * | 2020-09-02 | 2020-12-04 | 中国银行股份有限公司 | Bank customer satisfaction calculation method and device |
| CN112036328B (en) * | 2020-09-02 | 2024-02-27 | 中国银行股份有限公司 | Bank customer satisfaction calculating method and device |
| CN112183417A (en) * | 2020-09-30 | 2021-01-05 | 重庆天智慧启科技有限公司 | Business consultant service capability evaluation system and method |
| CN112200051A (en) * | 2020-09-30 | 2021-01-08 | 重庆天智慧启科技有限公司 | Case field inspection system and method |
| CN112200051B (en) * | 2020-09-30 | 2023-09-29 | 重庆天智慧启科技有限公司 | Case field inspection system and method |
| CN112183417B (en) * | 2020-09-30 | 2023-12-05 | 重庆天智慧启科技有限公司 | System and method for evaluating service capability of consultant in department of industry |
| CN112581230A (en) * | 2020-12-24 | 2021-03-30 | 安徽航天信息科技有限公司 | Commodity recommendation method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107341688A (en) | The acquisition method and system of a kind of customer experience | |
| CN108090433B (en) | Face recognition method and device, storage medium and processor | |
| CN108520226B (en) | Pedestrian re-identification method based on body decomposition and significance detection | |
| CN105389593B (en) | Image object recognition methods based on SURF feature | |
| CN103325122B (en) | Based on the pedestrian retrieval method of Bidirectional sort | |
| TWI687159B (en) | Fry counting system and fry counting method | |
| CN104408449B (en) | Intelligent mobile terminal scene literal processing method | |
| CN109101938B (en) | Multi-label age estimation method based on convolutional neural network | |
| CN108171184A (en) | Method for distinguishing is known based on Siamese networks again for pedestrian | |
| CN106529499A (en) | Fourier descriptor and gait energy image fusion feature-based gait identification method | |
| WO2016190814A1 (en) | Method and system for facial recognition | |
| CN108647625A (en) | A kind of expression recognition method and device | |
| CN105005765A (en) | Facial expression identification method based on Gabor wavelet and gray-level co-occurrence matrix | |
| CN103034838A (en) | Special vehicle instrument type identification and calibration method based on image characteristics | |
| CN105224921A (en) | A kind of facial image preferentially system and disposal route | |
| CN109740572A (en) | A kind of human face in-vivo detection method based on partial color textural characteristics | |
| CN111209818A (en) | Video individual identification method, system, equipment and readable storage medium | |
| CN105956570B (en) | Smiling face's recognition methods based on lip feature and deep learning | |
| CN108197636A (en) | A kind of paddy detection and sorting technique based on depth multiple views feature | |
| CN107392151A (en) | Face image various dimensions emotion judgement system and method based on neutral net | |
| CN103177266A (en) | Intelligent stock pest identification system | |
| CN109544523A (en) | Quality of human face image evaluation method and device based on more attribute face alignments | |
| CN106874825A (en) | The training method of Face datection, detection method and device | |
| CN107992783A (en) | Face image processing process and device | |
| CN110458792A (en) | Method and device for evaluating quality of face image |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171110 |
|
| RJ01 | Rejection of invention patent application after publication |