CN108711152A - A kind of image analysis method, device and user terminal based on AI technologies - Google Patents
A kind of image analysis method, device and user terminal based on AI technologies Download PDFInfo
- Publication number
- CN108711152A CN108711152A CN201810500163.4A CN201810500163A CN108711152A CN 108711152 A CN108711152 A CN 108711152A CN 201810500163 A CN201810500163 A CN 201810500163A CN 108711152 A CN108711152 A CN 108711152A
- Authority
- CN
- China
- Prior art keywords
- environment
- image
- recognition result
- technologies
- ambient image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/001—Industrial image inspection using an image reference approach
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of image analysis method, device and user terminal based on AI technologies, wherein the method includes：Obtain ambient image；AI identifications are carried out to the ambient image, generate recognition result；The recognition result is matched with the environment comparison information in environmental analysis library, matching result is generated, in order to position the fault zone of the ambient image according to the matching result.AI image recognitions are carried out to which failure judgement region improves the working efficiency of trouble hunting by the ambient image to acquisition, shorten the repair time, it avoids the missing inspection manually overhauled or the case where row can not check, ensure that the safe operation of rail traffic vehicles to a certain extent.
Description
Technical field
The present invention relates to field of artificial intelligence, more specifically to a kind of image analysis side based on AI technologies
Method, device and user terminal.
Background technology
Rail traffic refers to that vehicle in use needs a kind of vehicles travelled in certain tracks or transportation system.Most allusion quotation
The rail traffic of type is exactly the railway system being made of traditional train and standard railroad.It is polynary with train and railway technology
Change development, rail traffic shows more and more types, the land transport of long range is not only dispersed throughout, in being also widely used in
In short-range urban public transport.
Common rail traffic failure includes vehicle trouble and orbital facilities failure, for common rail traffic faults into
When row maintenance, it includes manually that vehicle and orbital facilities carry out system to all environment of track by naked eyes to be required to technical staff
It investigates one by one, subjective factor is larger, and the time, long efficiency was low, and the work for the trouble hunting investigation of service personnel brings inconvenience,
Easily there is missing inspection or row the case where can not check in artificial inspection, for rail traffic vehicles safe operation bring it is huge
Security risk.
Invention content
In view of this, the present invention provides a kind of image analysis method, device and user terminal based on AI technologies to solve
The deficiencies in the prior art.
To solve the above problems, the present invention provides a kind of image analysis method based on AI technologies, including：
Obtain ambient image；
AI identifications are carried out to the ambient image, generate recognition result；
The recognition result is matched with the environment comparison information in environmental analysis library, generates matching result, so as to
In the fault zone for positioning the ambient image according to the matching result.
Preferably, described " AI identifications being carried out to the ambient image, generate recognition result ", including：
Determine the recognizable object environment in the ambient image；
Obtain the corresponding location information of the recognizable object environment；
According to the location information, the sectional drawing of the corresponding recognizable object environment of the location information is obtained；
Generate recognition result.
Preferably, described " determining the recognizable object environment in the ambient image " includes：
Object detection frame based on deep learning analyzes the ambient image, obtains in the ambient image
Recognizable object environment.
Preferably, before described " recognition result is matched with the environment comparison information in environmental analysis library ",
Further include：
Extract sectional drawing corresponding with the recognizable object environment and location information in the recognition result；
Relative position information of the sectional drawing in affiliated ambient image is obtained according to the location information；
Based on the environmental analysis library, according to the sectional drawing and the relative position information judge it is corresponding described in can
Identify whether the relative position of target environment is correct；
If so, the recognition result is stored, in order to execute " by the environment in the recognition result and environmental analysis library
Comparison information is matched ".
Preferably, described " the corresponding recognizable mesh to be judged according to the sectional drawing and the relative position information
Whether the relative position for marking environment is correct " after, further include：
If it is not, then generating the warning message of relative position mistake corresponding with the recognizable object environment.
Preferably, the environment comparison information includes normal environment setting data and failure environment setting data；
Described " recognition result is matched with the environment comparison information in environmental analysis library, generates matching result,
In order to determine the fault zone of the ambient image according to the matching result ", including：
The recognition result is compared with the normal environment setting data in the environment comparison information；
If the recognition result and the normal environment setting comparing do not pass through, by the recognition result with it is described
Failure environment setting data are compared；
If the recognition result passes through with failure environment setting comparing, matching result is generated, in order to root
The fault zone of the ambient image is positioned according to the matching result.
Preferably, after described " recognition result is compared with failure environment setting data ", further include：
If the recognition result does not pass through with failure environment setting comparing, prompt to the recognizable object
The corresponding ambient image of environment is reacquired, and returns to " obtaining ambient image ".
In addition, to solve the above problems, the present invention also provides a kind of image analysis apparatus based on AI technologies, including：It connects
Receive module, identification module and matching module；
The receiving module, for obtaining ambient image；
The identification module generates recognition result for carrying out AI identifications to the ambient image；
The matching module, for the recognition result to be matched with the environment comparison information in environmental analysis library,
Matching result is generated, in order to position the fault zone of the ambient image according to the matching result.
In addition, to solve the above problems, the present invention also provides a kind of user terminal, including memory and processor, institute
Memory is stated for storing the image analysis program based on AI technologies, the processor operation image based on AI technologies point
Program is analysed so that the user terminal executes the image analysis method as described above based on AI technologies.
In addition, to solve the above problems, the present invention also provides a kind of computer readable storage medium, it is described computer-readable
The image analysis program based on AI technologies is stored on storage medium, the image analysis program based on AI technologies is by processor
The image analysis method as described above based on AI technologies is realized when execution.
A kind of image analysis method, device and user terminal based on AI technologies provided by the invention.Wherein, institute of the present invention
The method of offer further carries out AI identifications by obtaining ambient image to ambient image, then carries out to recognition result and ring
Comparison information in the analysis library of border is compared, to orient the fault zone in ambient image.Pass through the environment to acquisition
Image carries out AI image recognitions to which failure judgement region improves the working efficiency of trouble hunting, shortens the repair time, keeps away
Exempt from the missing inspection of artificial maintenance or the case where row can not check, ensure that the safe operation of rail traffic vehicles to a certain extent.
Description of the drawings
Fig. 1 is the structure for the hardware running environment being related to the present invention is based on the image analysis method example scheme of AI technologies
Schematic diagram；
Fig. 2 is that the present invention is based on the flow diagrams of the image analysis method first embodiment of AI technologies；
Fig. 3 is that the present invention is based on the flow diagrams of the image analysis method second embodiment of AI technologies；
Fig. 4 is that the present invention is based on the flow diagrams of the image analysis method 3rd embodiment of AI technologies；
Fig. 5 is that the present invention is based on the flow diagrams of the image analysis method fourth embodiment of AI technologies；
Fig. 6 is that the present invention is based on the flow diagrams of the 5th embodiment of image analysis method of AI technologies；
Fig. 7 is that the present invention is based on the high-level schematic functional block diagrams of the image analysis apparatus of AI technologies.
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific implementation mode
The embodiment of the present invention is described below in detail, in which the same or similar labels are throughly indicated same or like
Element or element with the same or similar functions.
In addition, term " first ", " second " are used for description purposes only, it is not understood to indicate or imply relative importance
Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or
Implicitly include one or more this feature.In the description of the present invention, the meaning of " plurality " is two or more,
Unless otherwise specifically defined.
In the present invention unless specifically defined or limited otherwise, term " installation ", " connected ", " connection ", " fixation " etc.
Term shall be understood in a broad sense, for example, it may be being fixedly connected, may be a detachable connection, or integral；Can be that machinery connects
It connects, can also be electrical connection；It can be directly connected, can also can be indirectly connected through an intermediary in two elements
The interaction relationship of the connection in portion or two elements.It for the ordinary skill in the art, can be according to specific feelings
Condition understands the concrete meaning of above-mentioned term in the present invention.
It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not intended to limit the present invention.
As shown in Figure 1, the structural schematic diagram of the hardware running environment for the terminal that Fig. 1, which is the embodiment of the present invention, to be related to.
Terminal of the embodiment of the present invention can be PC, can also be that smart mobile phone, tablet computer, E-book reader, MP3 are broadcast
Putting device, MP4 players, pocket computer etc. has the packaged type terminal device of display function.
As shown in Figure 1, the terminal may include：Processor 1001, such as CPU, network interface 1004, user interface
1003, memory 1005, communication bus 1002.Wherein, communication bus 1002 is for realizing the connection communication between these components.
User interface 1003 may include display screen, input unit such as keyboard, remote controler, and optional user interface 1003 can also include
Standard wireline interface and wireless interface.Network interface 1004 may include optionally standard wireline interface and wireless interface (such as
WI-FI interfaces).Memory 1005 can be high-speed RAM memory, can also be stable memory, such as magnetic disk storage.
Memory 1005 optionally can also be the storage device independently of aforementioned processor 1001.
Optionally, terminal can also include camera, RF (Radio Frequency, radio frequency) circuit, sensor, audio
Circuit, WiFi module etc..In addition, mobile terminal can also configure gyroscope, barometer, hygrometer, thermometer, infrared ray sensing
The other sensors such as device, details are not described herein.
It will be understood by those skilled in the art that the restriction of the not structure paired terminal of terminal shown in Fig. 1, may include ratio
More or fewer components are illustrated, certain components or different components arrangement are either combined.
As shown in Figure 1, as may include operating system, number in a kind of memory 1005 of computer readable storage medium
According to interface control program, network attachment procedure and image analysis program based on AI technologies.
A kind of image analysis method, device and user terminal based on AI technologies provided by the invention.Wherein, the method
AI image recognitions are carried out to which failure judgement region improves the working efficiency of trouble hunting, contracting by the ambient image to acquisition
The short repair time avoids the missing inspection manually overhauled or the case where row can not check, ensure that rail traffic to a certain extent
The safe operation of vehicle.
Embodiment 1：
With reference to Fig. 2, first embodiment of the invention provides a kind of image analysis method based on AI technologies, including：
Step S100 obtains ambient image；
A kind of image analysis method based on AI technologies above-mentioned, that the present embodiment is provided can be applied to terminal and clothes
It is engaged between end.As, the shooting, collecting to ambient image is carried out by service technician's handheld terminal, and then is adopted captured
The ambient image of collection is back to server-side.In addition it is also possible to for the all-in-one machine of acquisition and discriminance analysis, as, service technique people
Member holds the all-in-one machine, carries out carrying out the work such as comprehensive image acquisition, data analysis to rail traffic.The acquisition of image walks
Rapid may include the shooting, acquisition, transmission of image.
Above-mentioned, the ambient image may include in the present embodiment vehicle general image, vehicle spare and accessory parts layout
Picture, intra-locomotive circuit image, vehicle axles wheel hub setting image, hydraulic-pneumatic instrument setting image, driver operation environment map
Picture, orbital direction setting image, interorbital connection image etc.；In addition it is also possible to including but not limited to track related facility
Image, such as track peripheral base station setting image, track signal lamp setting image etc..
Step S200 carries out AI identifications to the ambient image, generates recognition result；
It is above-mentioned, AI identifications are carried out to ambient image, are identified by AI, to generate an identification knot to the ambient image
Fruit, to carry out further automatic investigation work.
Above-mentioned, AI identifications are also referred to as intelligent image identification technology, and image recognition technology is an important neck of artificial intelligence
Domain.It refers to carrying out Object identifying to image, to identify the target of various different modes and to the technology of picture.Image recognition technology
May be based on the main feature of image.Each image has its feature, and such as letter A has a point, and P, which has, to be enclosed and Y
Center have a acute angle etc..When to image recognition eye movement studies have shown that sight always concentrates in the main feature of image,
The place of image outline curvature maximum or contour direction suddenly change is exactly concentrated on, the information content in these places are maximum.And
The scanning route of eyes is also always gone to from a feature in another feature successively.It can be seen that in image recognition processes,
Perceptual mechanism need to exclude the redundant information of input, extract crucial information out.Meanwhile it must believe there are one integration is responsible in brain
The mechanism of breath, it can be the finish message obtained stage by stage at a complete consciousness image.
Above-mentioned, recognition result is the result information generated by AI identification technologies corresponding with ambient image, may include
For the information such as particular elements setting, arrangement, connection relation, shape or result in ambient image.
Step S300 matches the recognition result with the environment comparison information in environmental analysis library, generates matching
As a result, in order to position the fault zone of the ambient image according to the matching result.
Above-mentioned, environmental analysis library is preset carrying out the database of further intelligent Matching, wherein the analysis library
In may include normal environment setting information or picture, or the setting information or picture of failure environment, by by institute
The comparison of obtained recognition result and the relevant information in the library, so that it is determined that whether there is failure in the ambient image acquired
Region then can further position fault zone if there is fault zone, and then service personnel or backstage can be prompted to grasp
The investigation to specifying region or repair are carried out as personnel.
The method that the present embodiment is provided further carries out AI identifications by obtaining ambient image to ambient image, then
Recognition result is compared with the comparison information in environmental analysis library, to orient the faulty section in ambient image
Domain.AI image recognitions are carried out to which the work that failure judgement region improves trouble hunting is imitated by the ambient image to acquisition
Rate shortens the repair time, avoids the missing inspection manually overhauled or the case where row can not check, ensure that track to a certain extent
The safe operation of vehicular traffic.
Embodiment 2：
With reference to Fig. 3, second embodiment of the invention provides a kind of image analysis method based on AI technologies, is based on above-mentioned Fig. 2
Shown in first embodiment, the step S200 " carries out AI identifications, generate recognition result " to the ambient image, including：
Step S210 determines the recognizable object environment in the ambient image；
It is above-mentioned, when carrying out AI identifications, need to carry out to carrying out data screening or data cleansing in ambient image, to carry
Take out recognizable object environment therein.Wherein it is possible to include but not limited to the Automatic Optimal for acquired ambient image
The comparison finding step of step and preset data and etc..
It is above-mentioned, Automatic Optimal step be service personnel when carrying out malfunction elimination, to the figure of car body or track running environment
When as obtaining, it is understood that there may be backlight, dim light, shake diplopia, discoloration, torsional deformation etc. shooting problem, so as to cause obtaining
The ambient image serious distortion got, and then the further identification of influence, so needing to carry out certainly the ambient image got
Dynamic optimization, Automatic Optimal can include but is not limited to improve contrast, adjust automatically bright and dark light, automatically process diplopia, is automatic
Adjust the functions such as torsional deformation.Above-mentioned Optimization Steps, or when obtaining picture by service personnel or system maintenance personnel
It is operated manually and selects to carry out the optimization to ambient image.
Above-mentioned, the comparison with preset data is searched, to be compared by preset maintenance target data, to really
Determine in image with the presence or absence of maintenance target.For example, service personnel needs overhaul train console, adopted by shooting
Collect the ambient image of console, by being compared to the ambient image and preset maintenance target, search in image is system
No there are the images such as direction-control apparatus, pneumatic-hydraulic control device, train heat management system control device, and if so, will
The images such as direction-control apparatus, pneumatic-hydraulic control device, train heat management system control device in image are as in image
Recognizable object environment.
Step S220 obtains the corresponding location information of the recognizable object environment；
It is above-mentioned, in the present embodiment, recognizable object environment can be positioned by AI identification technologies.Wherein, it positions
Can be to generate frame of different shapes according to shape, for example, polygon, rectangle, circle etc..Wherein, location information, can be with
The pixel coordinate for recognizable object environment in integrated environment image including generation can also include the frame each generated
Relative position, or may include the area of the frame, size, by obtaining relevant ranging data to calculate correlation
Equipment, the specific size information of facility.
Step S230 obtains the corresponding recognizable object environment of the location information according to the location information
Sectional drawing；
The frame to each recognizable object environment that is above-mentioned, being obtained according to location information, and then according to the frame to described
Recognizable object environment carries out sectional drawing, to obtain the sectional drawing of each recognizable object environment in ambient image.
Step S240 generates recognition result.
It is above-mentioned, the sectional drawing of the recognizable object environment in ambient image is can include but is not limited in recognition result, is positioned
The data such as information, specific size size, shape, in order to further carry out malfunction elimination according to recognition result.
Embodiment 3：
With reference to Fig. 4, third embodiment of the invention provides a kind of image analysis method based on AI technologies, is based on above-mentioned Fig. 3
Shown in second embodiment, the step S210 " determining the recognizable object environment in the ambient image ", including：
Step S211, the object detection frame based on deep learning analyze the ambient image, obtain the ring
Recognizable object environment in the image of border.
It is above-mentioned, it is to be understood that being the popular research direction of comparison, it experienced traditional artificial object detection one
The frame of design feature+shallow-layer grader arrives the object detection frame of the End-To-End based on big data and deep neural network
The detection and positioning of the object in image are mainly realized in the development of frame by object detection frame.Object detection frame can wrap
Include but be not limited to CNN, YOLO, RCNN, SSD, NMS etc., used detection framework is YOLO technologies in the present embodiment.
Above-mentioned, the object detection frame needs to carry out the deep learning for target, by deep learning, to realize
Ambient image is analyzed, to find out the recognizable object environment in ambient image, and then positioned, outline the mesh
The operations such as mark environment, the sectional drawing for extracting corresponding target environment.
Embodiment 4：
With reference to Fig. 5, fourth embodiment of the invention provides a kind of image analysis method based on AI technologies, is based on above-mentioned Fig. 4
Shown in 3rd embodiment, the step S300 " carries out the environment comparison information in the recognition result and environmental analysis library
Before matching ", further include：
Step S400 extracts sectional drawing corresponding with the recognizable object environment and location information in the recognition result；
Step S500 obtains relative position information of the sectional drawing in affiliated ambient image according to the location information；
It is above-mentioned, before being matched, extract the sectional drawing and location information of relevant recognizable object environment.And by fixed
Relative position information of the position acquisition of information in ambient image.The relative position information can be recognizable object environment and its
The relative position of his associated components or facility.For example, component in the setting of different location, screw in wheel transmission
Setting etc..
Step S600 is based on the environmental analysis library, and it is right with it to be judged according to the sectional drawing and the relative position information
Whether the relative position for the recognizable object environment answered is correct.
Step S700, if so, the recognition result is stored, in order to execute " by the recognition result and environmental analysis
Environment comparison information in library is matched ".
Step S800, if it is not, then generating the warning message of relative position mistake corresponding with the recognizable object environment.
It is above-mentioned, may include the relative position information of different recognizable object environment in environmental analysis library, it is as different
The correct of facility, equipment etc. put and facilities.For example, train wheel drive link should connect driving wheel and driven wheel, from
And carry out power output；After obtaining ambient image, drive link appears in other positions, does not connect with driving wheel and driven wheel
It connects, then can determine that the drive link deviates there may be installation site or drops to other positions, then generate and the recognizable mesh
Mark the warning message of the corresponding relative position mistake of environment.If after the ambient image identification of acquisition, recognizable object is judged
The relative position of environment is correct, then can store the recognition result, in order to further to the analysis of recognition result and judgement.
Embodiment 5：
With reference to Fig. 6, fifth embodiment of the invention provides a kind of image analysis method based on AI technologies, is based on above-mentioned Fig. 2
Shown in first embodiment, the environment comparison information include normal environment setting data and failure environment setting data；It is described
Step S300 " recognition result is matched with the environment comparison information in environmental analysis library, generates matching result, so as to
In the fault zone for determining the ambient image according to the matching result ", including：
Step S310 carries out the normal environment setting data in the recognition result and the environment comparison information
It compares；
It is above-mentioned, when carrying out the analysis to recognition result, need into be about to recognition result and normal environment be arranged data into
Row compares.
Above-mentioned, it includes the normal shape with maintenance target facility, equipment, component that normal environment setting data, which are preset,
The relevant information of state, normal arrangement, normal facilities.It may include pictorial information, relative position information, put information, set
Confidence breath etc..
Step S320, if the recognition result does not pass through with normal environment setting comparing, by the identification
As a result it is compared with failure environment setting data；
It is above-mentioned, if recognition result does not pass through with normal environment comparing, judge that the recognizable object environment may
There are failure problems, so by the identification target environment and failure environment data carry out it is secondary compare, judge that the recognition result is
The no relevant data matches with failure environment data.Pass through in addition, if comparing, then judges that the recognizable object environment is just
The standing state set, is arranged or puts layout, does not occur dependent failure.
Step S330 generates matching result if the recognition result passes through with failure environment setting comparing,
In order to position the fault zone of the ambient image according to the matching result.
It is above-mentioned, if carried out after the comparison of data is arranged for recognition result and failure environment, by comparing, then generation
With as a result, the matching result can include but is not limited to relevant comparison information, the positioning letter of the recognizable object environment of failure
Breath, relative position information etc., in order to the positioning further for the recognizable object environment, and can notify or prompt phase
Operating personnel are closed, such as service personnel carries out the maintenance to the target.
Step S340 is prompted if the recognition result does not pass through with failure environment setting comparing to described
The corresponding ambient image of recognizable object environment is reacquired, and returns to " obtaining ambient image ".
It is above-mentioned, if the comparison for recognition result and failure environment being arranged data does not pass through, judge that the image may be deposited
In relevant issues, leads to not comparison, such as shooting problem etc. that data are set by failure environment, need to carry out to the position
Ambient image reacquisition, and then staff is prompted to carry out reacquisition shooting to the ambient image, so again into
Row obtains the step of ambient image, reanalyses judgement.
In addition, with reference to Fig. 7, the present invention also provides a kind of image analysis apparatus based on AI technologies, including：Receiving module
10, identification module 20 and matching module 30；
The receiving module 10, for obtaining ambient image；
The identification module 20 generates recognition result for carrying out AI identifications to the ambient image；
The matching module 30 is used for the environment comparison information progress in the recognition result and environmental analysis library
Match, matching result is generated, in order to position the fault zone of the ambient image according to the matching result.
In addition, the present invention also provides a kind of user terminal, including memory and processor, the memory is for storing
Based on the image analysis program of AI technologies, the processor operation image analysis program based on AI technologies is so that the use
Family terminal executes the image analysis method as described above based on AI technologies.
In addition, the present invention also provides a kind of computer readable storage medium, stored on the computer readable storage medium
There is the image analysis program based on AI technologies, is realized when the image analysis program based on AI technologies is executed by processor as above
State the image analysis method based on AI technologies.
It should be noted that herein, the terms "include", "comprise" or its any other variant are intended to non-row
His property includes, so that process, method, article or system including a series of elements include not only those elements, and
And further include other elements that are not explicitly listed, or further include for this process, method, article or system institute it is intrinsic
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including this
There is also other identical elements in the process of element, method, article or system.
The embodiments of the present invention are for illustration only, can not represent the quality of embodiment.
Through the above description of the embodiments, those skilled in the art can be understood that above-described embodiment side
Method can add the mode of required general hardware platform to realize by software, naturally it is also possible to by hardware, but in many cases
The former is more preferably embodiment.Based on this understanding, technical scheme of the present invention substantially in other words does the prior art
Going out the part of contribution can be expressed in the form of software products, which is stored in one as described above
In storage medium (such as ROM/RAM, magnetic disc, CD), including some instructions use so that a station terminal equipment (can be mobile phone,
Computer, server or network equipment etc.) execute method described in each embodiment of the present invention.
It these are only the preferred embodiment of the present invention, be not intended to limit the scope of the invention, it is every to utilize this hair
Equivalent structure or equivalent flow shift made by bright specification and accompanying drawing content is applied directly or indirectly in other relevant skills
Art field, is included within the scope of the present invention.
Claims (10)
1. a kind of image analysis method based on AI technologies, which is characterized in that including：
Obtain ambient image；
AI identifications are carried out to the ambient image, generate recognition result；
The recognition result is matched with the environment comparison information in environmental analysis library, matching result is generated, in order to root
The fault zone of the ambient image is positioned according to the matching result.
2. the image analysis method as described in claim 1 based on AI technologies, which is characterized in that described " to the ambient image
AI identifications are carried out, recognition result is generated ", including：
Determine the recognizable object environment in the ambient image；
Obtain the corresponding location information of the recognizable object environment；
According to the location information, the sectional drawing of the corresponding recognizable object environment of the location information is obtained；
Generate recognition result.
3. the image analysis method as claimed in claim 2 based on AI technologies, which is characterized in that described " to determine the environment map
As in recognizable object environment " include：
Object detection frame based on deep learning analyzes the ambient image, obtains knowing in the ambient image
Other target environment.
4. the image analysis method as claimed in claim 3 based on AI technologies, which is characterized in that described " by the recognition result
Matched with the environment comparison information in environmental analysis library " before, further include：
Extract sectional drawing corresponding with the recognizable object environment and location information in the recognition result；
Relative position information of the sectional drawing in affiliated ambient image is obtained according to the location information；
Based on the environmental analysis library, judged according to the sectional drawing and the relative position information corresponding described recognizable
Whether the relative position of target environment is correct；
If so, storing the recognition result, " recognition result and the environment in environmental analysis library are compared in order to execute
Information is matched ".
5. the image analysis method as claimed in claim 4 based on AI technologies, which is characterized in that it is described " according to the sectional drawing and
The relative position information judges whether the relative position of the corresponding recognizable object environment is correct " after, also wrap
It includes：
If it is not, then generating the warning message of relative position mistake corresponding with the recognizable object environment.
6. the image analysis method based on AI technologies as described in claim any one of 1-5, which is characterized in that
The environment comparison information includes normal environment setting data and failure environment setting data；
Described " recognition result is matched with the environment comparison information in environmental analysis library, generates matching result, so as to
In the fault zone for determining the ambient image according to the matching result ", including：
The recognition result is compared with the normal environment setting data in the environment comparison information；
If the recognition result does not pass through with normal environment setting comparing, by the recognition result and the failure
Environment setting data are compared；
If the recognition result passes through with failure environment setting comparing, matching result is generated, in order to according to institute
State the fault zone that matching result positions the ambient image.
7. the image analysis method as claimed in claim 6 based on AI technologies, which is characterized in that described " by the recognition result
It is compared with failure environment setting data " after, further include：
If the recognition result does not pass through with failure environment setting comparing, prompt to the recognizable object environment
Corresponding ambient image is reacquired, and returns to " obtaining ambient image ".
8. a kind of image analysis apparatus based on AI technologies, which is characterized in that including：Receiving module, identification module and matching mould
Block；
The receiving module, for obtaining ambient image；
The identification module generates recognition result for carrying out AI identifications to the ambient image；
The matching module is generated for matching the recognition result with the environment comparison information in environmental analysis library
Matching result, in order to position the fault zone of the ambient image according to the matching result.
9. a kind of user terminal, which is characterized in that including memory and processor, the memory is based on AI skills for storing
The image analysis program of art, the processor operation image analysis program based on AI technologies is so that the user terminal is held
Image analysis method of the row based on AI technologies as described in any one of claim 1-7.
10. a kind of computer readable storage medium, which is characterized in that be stored with based on AI on the computer readable storage medium
The image analysis program of technology realizes such as claim 1- when the image analysis program based on AI technologies is executed by processor
Image analysis method based on AI technologies described in any one of 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810500163.4A CN108711152B (en) | 2018-05-23 | 2018-05-23 | Image analysis method and device based on AI technology and user terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810500163.4A CN108711152B (en) | 2018-05-23 | 2018-05-23 | Image analysis method and device based on AI technology and user terminal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108711152A true CN108711152A (en) | 2018-10-26 |
| CN108711152B CN108711152B (en) | 2021-08-13 |
Family
ID=63869287
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810500163.4A Active CN108711152B (en) | 2018-05-23 | 2018-05-23 | Image analysis method and device based on AI technology and user terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108711152B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109685151A (en) * | 2018-12-27 | 2019-04-26 | 深圳市鹰眼在线电子科技有限公司 | The recognition methods of circuit board item number, device and computer readable storage medium |
| CN109784336A (en) * | 2019-01-30 | 2019-05-21 | 云南电网有限责任公司电力科学研究院 | A kind of infrared image fault point recognition methods based on YOLO algorithm of target detection |
| CN110263623A (en) * | 2019-05-07 | 2019-09-20 | 平安科技（深圳）有限公司 | Train climbs monitoring method, device, terminal and storage medium |
| CN113161113A (en) * | 2021-06-02 | 2021-07-23 | 国家电网有限公司 | Transformer substation air cooling visual management system and method |
| CN114187131A (en) * | 2021-12-16 | 2022-03-15 | 南京凯盛国际工程有限公司 | Cement factory safety management system based on intelligent glasses terminal |
| CN114248818A (en) * | 2022-01-13 | 2022-03-29 | 南京融才交通科技研究院有限公司 | Intelligent information transportation supervision method and system based on rail transit |
| CN115966018A (en) * | 2022-12-26 | 2023-04-14 | 北京思维实创科技有限公司 | Anomaly monitoring method based on machine vision and related device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101229232B1 (en) * | 2011-06-14 | 2013-02-04 | 한국철도기술연구원 | Detection apparatus and method of train outside fault |
| CN103512762A (en) * | 2012-06-29 | 2014-01-15 | 北京华兴致远科技发展有限公司 | Image processing method and device and train fault detection system |
| CN104200464A (en) * | 2014-08-19 | 2014-12-10 | 苏州华兴致远电子科技有限公司 | Train anomaly recognition detection method and system |
| CN104463235A (en) * | 2014-11-18 | 2015-03-25 | 中国铁道科学研究院电子计算技术研究所 | Fault recognition method and device based on operation images of motor train unit |
| CN107273802A (en) * | 2017-05-16 | 2017-10-20 | 武汉华目信息技术有限责任公司 | A kind of detection method and device of railroad train brake shoe drill ring failure |
| CN107392197A (en) * | 2017-07-27 | 2017-11-24 | 武汉汉云亚信科技有限公司 | A kind of locomotive identification identification and locomotive component fault detection method and system |
-
2018
- 2018-05-23 CN CN201810500163.4A patent/CN108711152B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101229232B1 (en) * | 2011-06-14 | 2013-02-04 | 한국철도기술연구원 | Detection apparatus and method of train outside fault |
| CN103512762A (en) * | 2012-06-29 | 2014-01-15 | 北京华兴致远科技发展有限公司 | Image processing method and device and train fault detection system |
| CN104200464A (en) * | 2014-08-19 | 2014-12-10 | 苏州华兴致远电子科技有限公司 | Train anomaly recognition detection method and system |
| CN104463235A (en) * | 2014-11-18 | 2015-03-25 | 中国铁道科学研究院电子计算技术研究所 | Fault recognition method and device based on operation images of motor train unit |
| CN107273802A (en) * | 2017-05-16 | 2017-10-20 | 武汉华目信息技术有限责任公司 | A kind of detection method and device of railroad train brake shoe drill ring failure |
| CN107392197A (en) * | 2017-07-27 | 2017-11-24 | 武汉汉云亚信科技有限公司 | A kind of locomotive identification identification and locomotive component fault detection method and system |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109685151A (en) * | 2018-12-27 | 2019-04-26 | 深圳市鹰眼在线电子科技有限公司 | The recognition methods of circuit board item number, device and computer readable storage medium |
| CN109685151B (en) * | 2018-12-27 | 2022-10-14 | 深圳市鹰眼在线电子科技有限公司 | Circuit board material number identification method and device and computer readable storage medium |
| CN109784336A (en) * | 2019-01-30 | 2019-05-21 | 云南电网有限责任公司电力科学研究院 | A kind of infrared image fault point recognition methods based on YOLO algorithm of target detection |
| CN110263623A (en) * | 2019-05-07 | 2019-09-20 | 平安科技（深圳）有限公司 | Train climbs monitoring method, device, terminal and storage medium |
| CN110263623B (en) * | 2019-05-07 | 2024-05-28 | 平安科技（深圳）有限公司 | Train climbing monitoring method, device, terminal and storage medium |
| CN113161113A (en) * | 2021-06-02 | 2021-07-23 | 国家电网有限公司 | Transformer substation air cooling visual management system and method |
| CN113161113B (en) * | 2021-06-02 | 2023-06-27 | 国家电网有限公司 | Transformer substation air cooling visual management system and method |
| CN114187131A (en) * | 2021-12-16 | 2022-03-15 | 南京凯盛国际工程有限公司 | Cement factory safety management system based on intelligent glasses terminal |
| CN114248818A (en) * | 2022-01-13 | 2022-03-29 | 南京融才交通科技研究院有限公司 | Intelligent information transportation supervision method and system based on rail transit |
| CN115966018A (en) * | 2022-12-26 | 2023-04-14 | 北京思维实创科技有限公司 | Anomaly monitoring method based on machine vision and related device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108711152B (en) | 2021-08-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108711152A (en) | A kind of image analysis method, device and user terminal based on AI technologies | |
| CN107782564B (en) | Automatic driving vehicle evaluation system and method | |
| CN103940620B (en) | Vehicle failure based on image recognition differentiates and the method and system of prompting | |
| CN108803604A (en) | Vehicular automatic driving method, apparatus and computer readable storage medium | |
| CN110516518A (en) | A kind of illegal manned detection method of non-motor vehicle, device and electronic equipment | |
| CN107948585A (en) | Video recording labeling method, device and computer-readable recording medium | |
| CN107977607A (en) | A kind of fatigue driving monitoring method based on machine vision | |
| CN108717711A (en) | A kind of rail traffic fault picture analysis method, device and system | |
| CN110928620A (en) | Method and system for evaluating distraction of automobile HMI design to attract driving attention | |
| CN104077568A (en) | High-accuracy driver behavior recognition and monitoring method and system | |
| US20130346060A1 (en) | Translation interfacing apparatus and method using vision tracking | |
| CN113423063A (en) | Vehicle monitoring method and device based on vehicle-mounted T-BOX, vehicle and medium | |
| CN112949457A (en) | Maintenance method, device and system based on augmented reality technology | |
| CN115082861A (en) | Personnel identity and safety violation identification method and system | |
| KR20220008492A (en) | Method for providing deep learning model based vehicle part testing service | |
| CN108376241A (en) | Rubbish pick-up method, rubbish pickup vehicle and computer readable storage medium | |
| CN105353508A (en) | Detection method of automobile liquid residues through adoption of smart glasses | |
| CN111179458A (en) | Method and device for routing inspection of rail transit equipment | |
| CN103279823B (en) | The detection method of vehicle axles overhaul process and system | |
| CN108881846A (en) | Information fusion method, device and computer readable storage medium | |
| CN111178181B (en) | Traffic scene segmentation method and related device | |
| CN115631002B (en) | Computer vision-based intelligent damage assessment method and system for vehicle insurance | |
| CN115438945A (en) | Risk identification method, device, equipment and medium based on power equipment inspection | |
| CN110956072B (en) | Driving skill training method based on big data analysis | |
| CN113887297A (en) | Safe driving monitoring method and device for forming data closed loop based on cloud |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |