CN110427962A - A kind of test method, electronic equipment and computer readable storage medium - Google Patents
A kind of test method, electronic equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN110427962A CN110427962A CN201910537640.9A CN201910537640A CN110427962A CN 110427962 A CN110427962 A CN 110427962A CN 201910537640 A CN201910537640 A CN 201910537640A CN 110427962 A CN110427962 A CN 110427962A
- Authority
- CN
- China
- Prior art keywords
- image
- test
- identification facility
- test image
- identification
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Prevention of errors by analysis, debugging or testing of software
- G06F11/3668—Testing of software
- G06F11/3672—Test management
- G06F11/3688—Test management for test execution, e.g. scheduling of test suites
-
- 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
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Computer Hardware Design (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Image Analysis (AREA)
Abstract
The present embodiments relate to data processing field, a kind of test method, electronic equipment and computer readable storage medium are disclosed.In the section Example of the application, which includes: to obtain test chart image set, and it includes test image that test image, which is concentrated,；Test chart image set is imported into transmitter；Wherein, transmitter is used to for the test image that test image is concentrated being transmitted separately to identification facility, and identification facility is used to be based on reference map image set, identify to from the received test image of transmitter；Identification facility is obtained to the recognition result of each test image；According to each recognition result, the test result of identification facility is determined.In the realization, the test of identification facility can be carried out automatically, reduce the testing time.
Description
Technical field
The present embodiments relate to data processing field, in particular to a kind of test method, electronic equipment and computer can
Read storage medium.
Background technique
The basic realization principle of face recognition software be by camera collection image, extract characteristic in image into
Pedestrian demonstrate,proves comparison.Wherein, the process of feature extraction is extremely important, this process may be by characteristic point position, distance, angle
The influence of the factors such as degree causes face that can not be identified, or judges by accident.It would therefore be desirable to which test should under mass data
Software performance, the index of this performance test mainly include misclassification rate (probability that other people are accidentally made to designated person), reject rate
(probability that designated person is accidentally made to other staff) and match time (from feature extraction to the time for obtaining recognition result).
However, it is found by the inventors that at least there are the following problems in the prior art: since we have no idea to recognition of face
Software carries out code revision, can only carry out Black-box Testing.Traditional test method is will first to take photo storage to recognition of face
In software, then allow tester stand the face recognition software camera before test.This process is relatively complicated, and
Because we need to carry out the test of mass data, this test method will need to do many repetitive works, can waste very
More times.
It should be noted that information is only used for reinforcing the reason to the background of the disclosure disclosed in above-mentioned background technology part
Solution, therefore may include the information not constituted to the prior art known to persons of ordinary skill in the art.
Summary of the invention
Embodiment of the present invention is designed to provide a kind of test method, electronic equipment and computer-readable storage medium
Matter makes it possible to carry out the test of identification facility automatically, reduces the testing time.
In order to solve the above technical problems, embodiments of the present invention provide a kind of test method, comprising the following steps: obtain
Test chart image set is taken, it includes test image that test image, which is concentrated,；Test chart image set is imported into transmitter；Wherein, transmitter is used for
The test image that test image is concentrated is transmitted separately to identification facility, identification facility is used to be based on reference map image set, to from biography
The defeated received test image of device is identified；Identification facility is obtained to the recognition result of each test image；According to each recognition result,
Determine the test result of identification facility.
Embodiments of the present invention additionally provide a kind of electronic equipment, comprising: at least one processor；And at least
The memory of one processor communication connection；Wherein, memory is stored with the instruction that can be executed by least one processor, instruction
It is executed by least one processor, so that at least one processor is able to carry out the test method referred to such as above embodiment.
Embodiments of the present invention additionally provide a kind of computer readable storage medium, are stored with computer program, calculate
The test method that above embodiment refers to is realized when machine program is executed by processor.
The test image collected is transmitted to identification in terms of existing technologies, by transmitter by embodiment of the present invention
Tool eliminates many repeatability without object being placed in the identification region of identification facility to obtain a large amount of test datas
Operation reduces the time for collecting test data, improves test speed.In addition to this, electronic equipment executes above-mentioned step automatically
Suddenly, compared to original test method, do not need artificially to control and artificial statistics, saved many times, save manpower at
This.
In addition, it includes N number of image data set that test image, which is concentrated,；Test image in same image data set included
Object is identical, and, it is distinct between the shooting feature of any two width test image in same image data set；N is positive whole
Number；It according to each recognition result, determines the test result of identification facility, specifically includes: for each image data set, according to identification
Tool determines that the performance based on image data set of identification facility refers to the recognition result of the test image in image data set
Mark；Wherein, the recognition result of test image includes the matching result of test image Yu each benchmark image, and/or, test image
Match time；Wherein, the matching result of test image and each benchmark image is used to determine reject rate, the misclassification rate in performance indicator
With any one or any combination in accuracy rate, the match time of test image is used to determine the Mean match in performance indicator
Time；According to the performance indicator based on each image data set of identification facility, the test result of identification facility is determined.The realization
In, it can more meticulously understand the performance of identification facility, and provide direction further to improve identification facility.
In addition, obtaining test chart image set, specifically include: obtaining original test image collection；According to original test image collection
The shooting feature of the object that test image is included and test image, classifies to original test image collection, obtains N number of image
Data group；By N number of image data set, test chart image set is formed.
In addition, determining the test result of identification facility according to each recognition result, specifically including: according to identification facility to each
The recognition result of test image determines the performance indicator of identification facility；Wherein, the recognition result of test image includes test image
With the matching result of each benchmark image, and/or, the match time of test image；Wherein, of test image and each benchmark image
It is used to determine any one in reject rate, misclassification rate and the accuracy rate in performance indicator or any combination, test chart with result
The match time of picture is used to determine the Mean match time in performance indicator.In the realization, test result can be automatically generated, is mentioned
High the degree of automation of electronic equipment, further reduces the testing time of identification facility.
In addition, determining reject rate according to the matching result of test image, specifically include: actually right according to each test image
The matching result of the benchmark image and test image and each benchmark image answered, determines the number of identification facility False Rejects；Root
According to the ratio of the intra-class testing number of the number and test chart image set of identification facility False Rejects, reject rate is determined；According to test
The matching result of image, determines misclassification rate, specifically includes: according to each test image actually corresponding benchmark image, and test
The matching result of image and each benchmark image determines the number that identification facility mistake receives；Received according to identification facility mistake
The ratio of testing time, determines misclassification rate between number and the class of test chart image set；According to the matching result of test image, determine quasi-
True rate, specifically includes: according to the matching of each test image actually corresponding benchmark image and test image and each benchmark image
As a result, determining that identification facility matches correct number；Total of correct number and identification facility is matched according to identification facility
With number, accuracy rate is determined.
In addition, before obtaining identification facility to the recognition result of each test image, test method further include: obtain benchmark
Image set；Reference map image set is imported into identification facility.In the realization, benchmark image can be automatically generated and import identification facility,
The degree of automation of electronic equipment is further improved, the testing time of identification facility is further reduced.
In addition, test image concentrates the test image including M object, M is positive integer；Reference map image set is obtained, specifically
Include: to be performed the following operation respectively for each object: being concentrated from test image, select T test image of each object, made
For the benchmark image of object, T is positive integer；By the benchmark image of each object, reference map image set is formed.
In addition, test method is executed by automation tools AutoIt.In the realization, the operation speed of the test method is improved
Degree.
In addition, the source of test image is internet or local storage space.In the realization, acquisition test image is shortened
Time.
Detailed description of the invention
One or more embodiments are illustrated by the picture in corresponding attached drawing, these exemplary theorys
The bright restriction not constituted to embodiment, the element in attached drawing with same reference numbers label are expressed as similar element, remove
Non- to have special statement, composition does not limit the figure in attached drawing.
Fig. 1 is the flow chart of the test method of first embodiment of the invention；
Fig. 2 is the flow chart of the test method of second embodiment of the present invention；
Fig. 3 is the schematic diagram for the test image for including in the image data set of second embodiment of the present invention；
Fig. 4 is the structural schematic diagram of the test device of third embodiment of the present invention；
Fig. 5 is the structural schematic diagram of the electronic equipment of the 4th embodiment of the invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with attached drawing to the present invention
Each embodiment be explained in detail.However, it will be understood by those skilled in the art that in each embodiment party of the present invention
In formula, in order to make the reader understand this application better, many technical details are proposed.But even if without these technical details
And various changes and modifications based on the following respective embodiments, the application technical solution claimed also may be implemented.
The first embodiment of the present invention is related to a kind of test methods, are applied to electronic equipment, which can be
The device of identification facility is installed, the test device with the device communication connection of installation identification facility is also possible to.The test side
Method is used to test the performance of identification facility.Wherein, identification facility can be face recognition software, article identification software, license plate and know
Other software etc., it is clear to illustrate, in present embodiment by taking face recognition software as an example, the test method of identification facility is lifted
Example explanation, it will be understood by those skilled in the art that the test method of other identification softwares can refer to the correlation of present embodiment
Content does not repeat one by one herein.Specifically, as shown in Figure 1, the test method includes:
Step 101: obtaining test chart image set.
Specifically, it includes test image that test image, which is concentrated, which can be to come from and collects and store in advance
In the image data of local storage space, it is also possible to what the downloading directly from internet obtained.
In one example, test chart image set is derived from online disclosed test database, and data sample is big and comprehensive, shortens
Tester collects the time of test image, and improves the accuracy of test result.
Step 102: test chart image set is imported into transmitter.
Specifically, transmitter is used to for the test image that test image is concentrated to be transmitted separately to identification facility, work is identified
Tool is identified for being based on reference map image set to from the received test image of transmitter.
In one example, transmitter is based on pre-set image transmitting mode, the test chart that test image is concentrated
As being transmitted separately to identification facility.Wherein, pre-set transmission mode can be the test image concentrated according to test image
Placement order, it is vertical that test image is transmitted to identification facility so that identification facility think the test image for know
The imaging sensor acquired image of other tool is divided into Q seconds between the transmission time of adjacent two width test image, the value of Q
It can according to need setting.
In one example, transmitter realizes that is, the transmitter can be with the above function based on virtual transmission device
Software.
It is noted that test image needs for object to be placed on the cog region of identification facility in original test method
Domain, carries out identification test, and in present embodiment, the test image collected is transmitted to identification facility by transmitter, without will
Object is placed in the identification region of identification facility to obtain a large amount of test datas, is eliminated many repetitive operations, is simplified
The input process of test data reduces the time of input test data, improves test speed.Wherein, the knowledge of identification facility
Other region refers to the shooting area of the imaging sensor of identification facility.
In one example, reference map image set is transmitted to identification facility by electronic equipment.Specifically, electronic equipment obtains
Take reference map image set；Reference map image set is imported into identification facility.Wherein, it includes benchmark image, identification facility that benchmark image, which is concentrated,
Characteristic point in benchmark image is extracted, with the characteristic point of each object of determination.
It is noted that electronic equipment automatically generates benchmark image and imports identification facility, electronics is further improved
The degree of automation of equipment further reduces the testing time of identification facility.
In one example, test image concentrates the test image including M object, and M is positive integer.Electronic equipment obtains
The method of reference map image set are as follows: electronic equipment is directed to each object, performs the following operation respectively: concentrating from test image, selection
T test image of object, as the benchmark image of object, T is positive integer；By the benchmark image of each object, benchmark is formed
Image set.For example, test image is facial image, T is equal to 1, and test image concentrates the facial image including M people, and electronics is set
It is standby to be concentrated from test image, select everyone 1 positive facial image, the benchmark image as people.
It is noted that since the characteristic point of the positive facial image of people is more than the face figure of the other angles of people
Picture enables identification facility more accurately to obtain the characteristic point in every face, reduces the characteristic point due to benchmark image
Deficiency leads to the situation of identification mistake.
It should be noted that it will be understood by those skilled in the art that also can choose the figure of other angles in practical application
As being used as benchmark image, present embodiment does not limit the type of the benchmark image of selection.
Step 103: obtaining identification facility to the recognition result of each test image.
Specifically, electronic equipment is after identification facility identifies test image, to the recognition result of identification facility
It is recorded, in order to which the performance to identification facility is analyzed.Alternatively, identification facility is after identifying test image,
Recognition result is recorded, electronic equipment obtains identification facility by way of transferring the recognition result of identification facility record
To the recognition result of each test image.
Step 104: according to each recognition result, determining the test result of identification facility.
Specifically, electronic equipment can be by identification facility to the recognition result of every width test image, as identification facility
Test result, feed back to party in request；Identification work can also be determined according to identification facility to the recognition result of every width test image
The performance indicator of tool, and the performance indicator based on identification facility, determine the test result of identification facility, feed back to party in request.
In one example, electronic equipment determines identification work according to identification facility to the recognition result of every width test image
The performance indicator of tool.Wherein, the recognition result of test image includes the matching result of test image Yu each benchmark image, and/or,
The match time of test image；Wherein, test image and the matching result of each benchmark image are refused for determining in performance indicator
Any one in knowledge rate, misclassification rate and accuracy rate or any combination, the match time of test image is for determining performance indicator
In Mean match time.Electronic equipment determines the test result of identification facility according to the performance indicator of identification facility.
It is noted that the performance indicator that electronic equipment automatically determines identification facility (includes at least misclassification rate, rejection
Any one in rate, accuracy rate and Mean match time), test result is generated, the degree of automation of electronic equipment is improved,
Further reduce the testing time of identification facility.
In one example, electronic equipment according to each test image actually corresponding benchmark image and test image with
The matching result of each benchmark image determines the number of identification facility False Rejects；According to the number of identification facility False Rejects and
The ratio of the intra-class testing number of test chart image set, determines reject rate.Electronic equipment is according to the practical corresponding base of each test image
The matching result of quasi- image and test image and each benchmark image determines the number that identification facility mistake receives；According to identification
The ratio of testing time, determines misclassification rate between the class of number and test chart image set that tool mistake receives.Electronic equipment is according to each
Test image actually corresponding benchmark image and the matching result of test image and each benchmark image, determine identification facility
With correct number；Total matching times that correct number and identification facility are matched according to identification facility, determine accuracy rate.
Wherein, False Rejects refer to: when by test image and test image, actually corresponding benchmark image is matched, identification facility
It should be judged to successful match, be but mistakenly judged to that it fails to match.Mistake receiving refers to: by test image and except test image is real
When other benchmark images other than the corresponding benchmark image in border are matched, identification facility should be judged to that it fails to match, but mistake
Ground is judged to successful match.Intra-class testing refers to that actually corresponding benchmark image is matched test image with test image, between class
Test refers to that test image is matched with the image in addition to the test image actually corresponding benchmark image.
Below in conjunction with actual scene, the process of the performance indicator of identification facility, which is illustrated, to be determined to electronic equipment.
It is assumed that there is 100 objects, each object has 10 pictures, and test chart image set shares 100*10=1000 picture.
Reference map image set (100) of the front picture as identification facility, identification facility difference are respectively extracted for 100 objects
It is matched based on the picture in test chart image set (1000) with reference map image set.
Enabling test image A, actually corresponding benchmark image is benchmark image A.If the matching of test image A and benchmark image A
As a result it fails to match for instruction, then the problem of False Rejects occurs in identification facility, if test image A and benchmark image A
Matching result indicates successful match, then recognition result matching is correct.If the matching result of test image A and benchmark image B indicate
Successful match, then there is the problem of primary mistake receives in identification facility, if the matching result of test image A and benchmark image B
It fails to match for instruction, then identification facility matching is correct.
In one example, electronic equipment can judge whether the two is real based on test image and the mark of benchmark image
Border is corresponding.For example, add identical mark in advance for all images of same object, electronic equipment if it is determined that test image mark
Know consistent with the mark of benchmark image, then both be same object, i.e., it is both practical corresponding, however, it is determined that the mark of test image with
The mark of benchmark image is inconsistent, it is determined that the two is not same object, i.e., the two is practical does not correspond to.
It can calculate separately to obtain reject rate (survey in the number and class of False Rejects based on above-mentioned matching result and recognition result
The total degree of examination), (identification is correctly for misclassification rate (number that mistake receives and the ratio of total degree tested between class) and accuracy rate
The ratio of number and total matching times), and then obtain the performance indicator of identification facility.
It should be noted that it will be understood by those skilled in the art that the performance indicator of identification facility may be used also in practical application
To include other indexs such as accuracy rate, be not listed one by one herein.
In one example, tester writes automatized script using computer programming language Python, so that
Electronic equipment arranges the performance indicator of identification facility, generates test report by executing the automatized script.
In one example, the test side that present embodiment refers to is executed by the automation tools AutoIt in electronic equipment
Method.For example, identification facility is face recognition software, and tester writes automatized script, by test chart by AutoIt
Image set is imported into transmitter, and opens face recognition software.
AutoIt is concentrated from test image, and the positive test image for extracting everyone imported into people as benchmark image
In face identification software, for being matched with test image.
AutoIt starts transmitter, and it is soft that the test image that test image is concentrated is transferred to recognition of face by transmitter a sheet by a sheet
In the data acquisition equipment of part, such as camera, data acquisition equipment by collected data be input to face recognition software into
Row identification.Face recognition software exports the recognition result to test image, wherein the recognition result of test image includes test chart
As the matching result with each benchmark image, and/or, the match time of test image.
AutoIt records result and this match time of face recognition software, and can carry out to the recognition result of record
Processing, including the benchmark image being matched to according to identification facility and the corresponding true benchmark image of the test image, certainty
Can any one or any combination in reject rate, misclassification rate and the accuracy rate in index, according to the match time of test image,
Determine the Mean match time in performance indicator；AutoIt is using python script by the performance indicator of face recognition software into one
Step is arranged, and test report, the test result as face recognition software are generated.
It should be noted that it will be understood by those skilled in the art that in present embodiment, with being embodied as based on AutoIt
Example, is illustrated the implementation process of test method, in practical application, can also pass through the automatic chemical industry of other computer software
Have to realize test method that present embodiment refers to.
It is noted that improving the test since Autoit can realize automatically each function by Run Script
The service speed of method.
It should be noted that the above is only limit for example, not constituting to technical solution of the present invention.
Compared with prior art, the test method provided in present embodiment, the test image that will have been collected by transmitter
It is transmitted to identification facility, it is not necessary that object to be placed in the identification region of identification facility to obtain a large amount of test datas, is eliminated
Many repetitive operations reduce the time for collecting test data, simplify the input process of test data, improve test speed
Degree.In addition to this, electronic equipment is based on automation tools, executes above-mentioned steps automatically, compared to original test method, saves
It many times, does not need artificially to control and artificial statistics, saves human cost.
Second embodiment of the present invention is related to a kind of test method.Present embodiment is to first embodiment into one
Step refinement, specifically illustrates: obtaining the mode of test chart image set, and under the mode of the acquisition test chart image set, test knot
The method of determination of fruit.
Specifically, as shown in Fig. 2, in the present embodiment, include step 201 to step 207, wherein step 204 and
Step 205 respectively in first embodiment step 102 and step 103 it is roughly the same, details are not described herein again.It is main below to be situated between
Continue difference:
Step 201: obtaining original test image collection.
Specifically, the test image concentrated of original test image can the disclosed identification facility from network test number
According to downloading in library, it is also possible to store in local storage space.
Step 202: the shooting feature of the object that the test image according to original test image collection is included and test image,
Classify to original test image collection, obtains N number of image data set.
Specifically, the object that the test image in same image data set is included is identical, and, same image data set
In any two width test image shooting feature between it is distinct.
In one example, the test image that original test image is concentrated is facial image, and shooting feature includes face
Shooting angle and facial expression.The facial image of the different shooting angles of the same person can be placed on an image by electronic equipment
In data group, the test image for including in the image data set is as shown in Figure 3.It can also be by the different facial expressions of the same person
Facial image is placed in an image data set.
It should be noted that it will be understood by those skilled in the art that shooting feature is also possible to other spies in practical application
Sign, for example, the coverage extent etc. when shooting brightness, shooting, is not listed one by one herein.
Step 203: by N number of image data set, forming test chart image set.
Specifically, electronic equipment is using the good original test image collection of taxonomic revision as test chart image set.
Step 204: test chart image set is imported into transmitter.
Step 205: obtaining identification facility to the recognition result of each test image.
Step 206: being directed to each image data set, the knowledge according to identification facility to the test image in the image data set
Not as a result, determining the performance indicator based on the image data set of identification facility.
Specifically, electronic equipment is directed to each image data set, the statistics of the performance indicator of identification facility is carried out.Its
In, electronic equipment determines the identification based on the image data set according to the recognition result of width test image every in image data set
The recognition result for the every width test image concentrated in the method and first embodiment of the performance indicator of tool according to test image,
The method for determining the performance indicator of identification facility is roughly the same, and details are not described herein again, and those skilled in the art can refer to first
The related content of embodiment executes step 206.
Step 207: according to the performance indicator based on each image data set of identification facility, determining the test of identification facility
As a result.
Specifically, electronic equipment can be directed to each image data set, by the spy of shooting corresponding to the image data set
Sign, with performance indicator of the identification facility after identifying the image data set, is recorded in test report correspondingly, will test
Report is used as test result.
For example, identification facility is face recognition software, identification image is facial image, and it includes 3 figures that test image, which is concentrated,
As data group.First image data set includes the facial image that first man is shot under different shooting angles, i.e. the first figure
As the corresponding shooting feature of data group is shooting angle.Second image data set includes that second people claps in different facial expressions
The corresponding shooting feature of the facial image taken the photograph, i.e. the second image data set is facial expression.Third image data set includes the
The facial image that three people shoot under different shooting brightness, the i.e. corresponding shooting feature of third image data set are shooting
Brightness.Electronic equipment, to the recognition result of the test image in the first image data set, determines the people according to face recognition software
The performance indicator based on the first image data set of face identification software.Electronic equipment is according to face recognition software to the second picture number
According to the recognition result of the test image in group, the performance indicator based on the second image data set of the face recognition software is determined.
Electronic equipment, to the recognition result of the test image in third image data set, determines the recognition of face according to face recognition software
The performance indicator based on third image data set of software.Electronic equipment in table form, records each image data set pair
The performance indicator based on image data set of the shooting feature and face recognition software answered, obtains test report, and by the survey
Examination report is used as test result.In the example, test result is as shown in table 1.
Table 1
| Shoot feature | Performance indicator |
| Shooting angle | Performance indicator based on the first image data set |
| Facial expression | Performance indicator based on the second image data set |
| Shoot brightness | Performance indicator based on third image data set |
It is noted that carrying out the test of identification facility for the test chart image set by taxonomic revision, can obtain
For the performance indicator of the identification facility of each shooting feature, and then property of the difference to identification facility of shooting feature can be considered
The influence degree of energy more meticulously understands the performance of identification facility, and provides direction further to improve identification facility.
In one example, electronic equipment determine identification facility for each image data set performance indicator after, according to
The recognition result of the test image of all image data sets, alternatively, the performance based on each image data set of identification facility refers to
Mark determines the performance indicator of the identification facility based on test chart image set.For example, by the property based on image data set of identification facility
The average value of energy index, the performance indicator of the identification facility based on test chart image set as identification facility.Wherein, based on test
The performance indicator of the identification facility of image set can be used for evaluating the overall performance of identification facility.
It should be noted that the above is only limit for example, not constituting to technical solution of the present invention.
Compared with prior art, the test method provided in present embodiment, electronic equipment can be special for each shooting
Sign carries out the statistics of the performance indicator of identification facility, understands the performance of identification facility, and more meticulously further to improve identification
Tool provides direction.
The step of various methods divide above, be intended merely to describe it is clear, when realization can be merged into a step or
Certain steps are split, multiple steps are decomposed into, as long as including identical logical relation, all in the protection scope of this patent
It is interior；To adding inessential modification in algorithm or in process or introducing inessential design, but its algorithm is not changed
Core design with process is all in the protection scope of the patent.
Third embodiment of the present invention is related to a kind of test device, as shown in Figure 4, comprising: the first acquisition module 401,
Import modul 402, second obtains module 403 and determining module 404.First acquisition module 401 is surveyed for obtaining test chart image set
Attempt to include test image in image set；Import modul 402 is used to test chart image set importing transmitter；Wherein, transmitter is used for
The test image that test image is concentrated is transmitted separately to identification facility, identification facility is used to be based on reference map image set, to from biography
The defeated received test image of device is identified；Second acquisition module 403 is for obtaining identification of the identification facility to each test image
As a result；Determining module 404 is used to determine the test result of identification facility according to each recognition result.
In one example, it includes N number of image data set that test image, which is concentrated,；Test image in same image data set
The object for being included is identical, and, it is distinct between the shooting feature of any two width test image in same image data set；
N is positive integer；It according to each recognition result, determines the test result of identification facility, specifically includes: being directed to each image data set,
According to identification facility to the recognition result of the test image in image data set, determine identification facility based on image data set
Performance indicator；Wherein, the recognition result of test image includes the matching result of test image Yu each benchmark image, and/or, test
The match time of image；Wherein, the matching result of test image and each benchmark image can be used for determining the rejection in performance indicator
Any one in rate, misclassification rate and accuracy rate or any combination, the match time of test image can be used for determining performance indicator
In Mean match time；According to the performance indicator based on each image data set of identification facility, the survey of identification facility is determined
Test result.
In one example, the first acquisition module 401 is specifically used for: obtaining original test image collection；According to original test
The shooting feature of the object that the test image of image set is included and test image, classifies to original test image collection, obtains
To N number of image data set；By N number of image data set, test chart image set is formed.
In one example, determining module 404 is specifically used for: according to identification facility to the identification knot of every width test image
Fruit determines the performance indicator of identification facility；Wherein, the recognition result of test image includes of test image Yu each benchmark image
With as a result, and/or, the match time of test image；Wherein, the matched benchmark image of identification facility can be used for determining performance indicator
In reject rate, any one or any combination, the match time of test image in misclassification rate and accuracy rate can be used for determining
Mean match time in performance indicator；According to each recognition result, the performance indicator of identification facility is determined.
In one example, test device further includes that third obtains module, and third obtains module and is used to obtain mould second
Before block 403 obtains identification facility to the recognition result of each test image, reference map image set is obtained；Reference map image set is imported and is known
Other tool.
In one example, test image concentrates the test image including M object, and M is positive integer；Obtain benchmark image
Collection, specifically includes: for each object, performing the following operation respectively: concentrating from test image, select T test chart of object
Picture, as the benchmark image of object, T is positive integer；By the benchmark image of each object, reference map image set is formed.
In one example, the source of test image is internet or local storage space.
It is not difficult to find that present embodiment is Installation practice corresponding with first embodiment, present embodiment can be with
First embodiment is worked in coordination implementation.The relevant technical details mentioned in first embodiment still have in the present embodiment
Effect, in order to reduce repetition, which is not described herein again.Correspondingly, the relevant technical details mentioned in present embodiment are also applicable in
In first embodiment.
It is noted that each module involved in present embodiment is logic module, and in practical applications, one
A logic unit can be a physical unit, be also possible to a part of a physical unit, can also be with multiple physics lists
The combination of member is realized.In addition, in order to protrude innovative part of the invention, it will not be with solution institute of the present invention in present embodiment
The technical issues of proposition, the less close unit of relationship introduced, but this does not indicate that there is no other single in present embodiment
Member.
4th embodiment of the invention is related to a kind of electronic equipment, as shown in Figure 5, comprising: at least one processor
501；And the memory 502 with the communication connection of at least one processor 501；Wherein, be stored with can be by least for memory 502
The instruction that one processor 501 executes, instruction is executed by least one processor 501, so that at least one processor 501 can
Execute the test method referred to such as above embodiment.
The electronic equipment includes: one or more processors 501 and memory 502, with a processor 501 in Fig. 5
For.Processor 501, memory 502 can be connected by bus or other modes, in Fig. 5 for being connected by bus.
Memory 502 is used as a kind of non-volatile computer readable storage medium storing program for executing, can be used for storing non-volatile software program, non-volatile
Property computer executable program and module.The non-volatile software journey that processor 501 is stored in memory 502 by operation
Sequence, instruction and module realize above-mentioned test method thereby executing the various function application and data processing of equipment.
Memory 502 may include storing program area and storage data area, wherein storing program area can store operation system
Application program required for system, at least one function；It storage data area can the Save option list etc..In addition, memory 502 can be with
It can also include nonvolatile memory, for example, at least disk memory, a flash memory including high-speed random access memory
Device or other non-volatile solid state memory parts.In some embodiments, it includes relative to processing that memory 502 is optional
The remotely located memory of device 501, these remote memories can pass through network connection to external equipment.The example of above-mentioned network
Including but not limited to internet, intranet, local area network, mobile radio communication and combinations thereof.
One or more module is stored in memory 502, when being executed by one or more processor 501, is held
Test method in the above-mentioned any means embodiment of row.
The said goods can be performed the application embodiment provided by method, have the corresponding functional module of execution method and
Beneficial effect, the not technical detail of detailed description in the present embodiment, reference can be made to method provided by the application embodiment.
5th embodiment of the invention is related to a kind of computer readable storage medium, is stored with computer program.It calculates
Machine program realizes above method embodiment when being executed by processor.
That is, it will be understood by those skilled in the art that implement the method for the above embodiments be can be with
Relevant hardware is instructed to complete by program, which is stored in a storage medium, including some instructions are to make
It obtains an equipment (can be single-chip microcontroller, chip etc.) or processor (processor) executes side described in each embodiment of the application
The all or part of the steps of method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only
Memory), random access memory (RAM, Random Access Memory), magnetic or disk etc. are various can store journey
The medium of sequence code.
It will be understood by those skilled in the art that the respective embodiments described above are to realize specific embodiments of the present invention,
And in practical applications, can to it, various changes can be made in the form and details, without departing from the spirit and scope of the present invention.
Claims (11)
1. a kind of test method characterized by comprising
Obtain test chart image set；
The test chart image set is imported into transmitter；Wherein, the transmitter is used for the test chart for concentrating the test image
As being transmitted separately to identification facility, the identification facility is used to be based on reference map image set, to from the received test of the transmitter
Image is identified；
The identification facility is obtained to the recognition result of each test image；
According to each recognition result, the test result of the identification facility is determined.
2. test method according to claim 1, which is characterized in that it includes N number of image data that the test image, which is concentrated,
Group；The object that test image in same described image data group is included is identical, and, appointing in same described image data group
It is distinct between the shooting feature of two width test images of anticipating；N is positive integer；
The recognition result according to the identification facility, determines the test result of the identification facility, specifically includes:
Identification for each described image data group, according to the identification facility to the test image in described image data group
As a result, determining the performance indicator based on described image data group of the identification facility；Wherein, the identification knot of the test image
Fruit includes the matching result of the test image Yu each benchmark image, and/or, the match time of the test image；Its
In, the matching result of the test image and the benchmark image is used to determine reject rate, the misclassification rate in the performance indicator
With any one or any combination in accuracy rate, the match time of the test image is for determining in the performance indicator
Mean match time；
According to the performance indicator based on each described image data group of the identification facility, the test of the identification facility is determined
As a result.
3. test method according to claim 2, which is characterized in that the acquisition test chart image set specifically includes:
Obtain original test image collection；
The shooting feature of object and the test image that test image according to the original test image collection is included, to institute
It states original test image collection to classify, obtains N number of described image data group；
By N number of image data set, the test chart image set is formed.
4. test method according to claim 1, which is characterized in that the recognition result according to the identification facility,
The test result for determining the identification facility, specifically includes:
According to the identification facility to the recognition result of each test image, the performance indicator of the identification facility is determined；Its
In, the recognition result of the test image includes the matching result of the test image Yu each benchmark image, and/or, institute
State the match time of test image Yu each benchmark image；Wherein, the matching of the test image and each benchmark image
As a result it is used to determine any one in reject rate, misclassification rate and the accuracy rate in the performance indicator or any combination, it is described
The match time of test image is used to determine the Mean match time in the performance indicator；
According to the performance indicator of the identification facility, the test result of the identification facility is determined.
5. test method according to claim 2 or 4, which is characterized in that according to the matching result of the test image, really
The fixed reject rate, specifically includes: according to each test image actually corresponding benchmark image and the test image with
The matching result of each benchmark image determines the number of the identification facility False Rejects；According to the identification facility mistake
The ratio of the intra-class testing number of the number of refusal and the test chart image set, determines the reject rate；
It according to the matching result of the test image, determines the misclassification rate, specifically includes: is practical according to each test image
The matching result of corresponding benchmark image and the test image and each benchmark image determines that the identification facility is wrong
The number that misconnection is received；The ratio of testing time between the class of the number and the test chart image set that are received according to the identification facility mistake
Value, determines the misclassification rate；
It according to the matching result of the test image, determines the accuracy rate, specifically includes: is practical according to each test image
The matching result of corresponding benchmark image and the test image and each benchmark image, determines the identification facility
With correct number；Total matching times of correct number and the identification facility are matched according to the identification facility, are determined
The accuracy rate.
6. test method according to claim 1, which is characterized in that obtain the identification facility to each survey described
Before the recognition result for attempting picture, the test method further include:
Obtain reference map image set；
The reference map image set is imported into the identification facility.
7. test method according to claim 6, which is characterized in that the test image concentrates the survey including M object
Attempt picture, M is positive integer；
It is described to obtain the reference map image set, it specifically includes:
It for each object, performs the following operation: being concentrated from the test image respectively, select T of each object
Test image, as the benchmark image of the object, T is positive integer；
By the benchmark image of each object, the reference map image set is formed.
8. test method according to any one of claim 1 to 7, which is characterized in that the test method is by automating
Tool AutoIt is executed.
9. test method according to any one of claim 1 to 7, which is characterized in that the source of the test image is
Internet or local storage space.
10. a kind of electronic equipment characterized by comprising at least one processor；And
The memory being connect at least one described processor communication；Wherein, be stored with can be by described at least one for the memory
The instruction that a processor executes, described instruction is executed by least one described processor, so that at least one described processor energy
Enough execute test method as claimed in any one of claims 1-9 wherein.
11. a kind of computer readable storage medium, is stored with computer program, which is characterized in that the computer program is located
Reason device realizes test method described in any one of claims 1 to 9 when executing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910537640.9A CN110427962A (en) | 2019-06-20 | 2019-06-20 | A kind of test method, electronic equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910537640.9A CN110427962A (en) | 2019-06-20 | 2019-06-20 | A kind of test method, electronic equipment and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110427962A true CN110427962A (en) | 2019-11-08 |
Family
ID=68408802
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910537640.9A Pending CN110427962A (en) | 2019-06-20 | 2019-06-20 | A kind of test method, electronic equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110427962A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826646A (en) * | 2019-11-27 | 2020-02-21 | 深圳市优必选科技股份有限公司 | Robot vision testing method, device, storage medium and terminal equipment |
| CN111209191A (en) * | 2019-12-29 | 2020-05-29 | 的卢技术有限公司 | An automated testing method and system for realizing video classification |
| CN112052820A (en) * | 2020-09-15 | 2020-12-08 | 展讯通信（上海）有限公司 | Test method and device for identifying AI scene identification rate, storage medium and terminal |
| CN112559332A (en) * | 2020-07-13 | 2021-03-26 | 支付宝(杭州)信息技术有限公司 | Attack testing method, device and equipment for biological feature recognition |
| WO2021217467A1 (en) * | 2020-04-28 | 2021-11-04 | 华为技术有限公司 | Method and apparatus for testing intelligent camera |
| CN115147700A (en) * | 2022-06-23 | 2022-10-04 | 中国电子技术标准化研究院 | An image recognition system target recognition rate parameter calibration method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106682676A (en) * | 2015-11-05 | 2017-05-17 | 中国移动通信集团天津有限公司 | Image test method and image test device |
| CN109145813A (en) * | 2018-08-20 | 2019-01-04 | 北京京东金融科技控股有限公司 | A kind of image matching algorithm test method and device |
| CN109376067A (en) * | 2018-08-15 | 2019-02-22 | 中国平安人寿保险股份有限公司 | Monkey test method, device, electronic equipment and computer readable storage medium |
| CN109766915A (en) * | 2018-12-14 | 2019-05-17 | 深圳壹账通智能科技有限公司 | Test method, device, computer equipment and storage medium based on image recognition |
-
2019
- 2019-06-20 CN CN201910537640.9A patent/CN110427962A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106682676A (en) * | 2015-11-05 | 2017-05-17 | 中国移动通信集团天津有限公司 | Image test method and image test device |
| CN109376067A (en) * | 2018-08-15 | 2019-02-22 | 中国平安人寿保险股份有限公司 | Monkey test method, device, electronic equipment and computer readable storage medium |
| CN109145813A (en) * | 2018-08-20 | 2019-01-04 | 北京京东金融科技控股有限公司 | A kind of image matching algorithm test method and device |
| CN109766915A (en) * | 2018-12-14 | 2019-05-17 | 深圳壹账通智能科技有限公司 | Test method, device, computer equipment and storage medium based on image recognition |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826646A (en) * | 2019-11-27 | 2020-02-21 | 深圳市优必选科技股份有限公司 | Robot vision testing method, device, storage medium and terminal equipment |
| CN111209191A (en) * | 2019-12-29 | 2020-05-29 | 的卢技术有限公司 | An automated testing method and system for realizing video classification |
| WO2021217467A1 (en) * | 2020-04-28 | 2021-11-04 | 华为技术有限公司 | Method and apparatus for testing intelligent camera |
| CN112559332A (en) * | 2020-07-13 | 2021-03-26 | 支付宝(杭州)信息技术有限公司 | Attack testing method, device and equipment for biological feature recognition |
| CN112052820A (en) * | 2020-09-15 | 2020-12-08 | 展讯通信（上海）有限公司 | Test method and device for identifying AI scene identification rate, storage medium and terminal |
| CN115147700A (en) * | 2022-06-23 | 2022-10-04 | 中国电子技术标准化研究院 | An image recognition system target recognition rate parameter calibration method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110427962A (en) | A kind of test method, electronic equipment and computer readable storage medium | |
| CN105184778B (en) | Detection method and device | |
| WO2022001097A1 (en) | Algorithm evaluation system and test method for performance test of person and certificate verification device | |
| CN110472494A (en) | Face feature extracts model training method, facial feature extraction method, device, equipment and storage medium | |
| CN104751110A (en) | Bio-assay detection method and device | |
| CN111738349B (en) | Detection effect evaluation method and device of target detection algorithm, storage medium and equipment | |
| CN109116129B (en) | Terminal detection method, detection device, system and storage medium | |
| CN110245573A (en) | A kind of register method, apparatus and terminal device based on recognition of face | |
| CN113033305B (en) | Living body detection method, living body detection device, terminal equipment and storage medium | |
| CN111583180A (en) | Image tampering identification method and device, computer equipment and storage medium | |
| CN106331524A (en) | Method and device for recognizing shot cut | |
| CN112560989B (en) | Artificial intelligence anti-counterfeiting image identification method and system based on big data | |
| CN118692155B (en) | Image detection method and device | |
| CN113627576B (en) | Code scanning information detection method, device, equipment and storage medium | |
| Ferrara et al. | Unsupervised fusion for forgery localization exploiting background information | |
| CN116630752B (en) | Construction site target object identification method and device based on AI algorithm | |
| CN111460419B (en) | Internet of things artificial intelligence face verification method and Internet of things cloud server | |
| CN113869364A (en) | Image processing method, image processing apparatus, electronic device, and medium | |
| You et al. | Tampering detection and localization base on sample guidance and individual camera device convolutional neural network features | |
| Bammey | Analysis and experimentation on the ManTraNet image forgery detector | |
| KR102482573B1 (en) | System and Method for tree species detection through tree bark Image background removal using deep learning | |
| US12272175B2 (en) | Evaluating method and system for face verification, and computer storage medium | |
| Cozzolino et al. | A comparative analysis of forgery detection algorithms | |
| CN117058440A (en) | Safety harness identification method, device, equipment and storage medium | |
| CN114092743B (en) | Compliance detection method and device for sensitive picture, storage medium and equipment |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191108 |