CN110390265A - A kind of recognition detection method and system of unmanned plane inspection - Google Patents
A kind of recognition detection method and system of unmanned plane inspection Download PDFInfo
- Publication number
- CN110390265A CN110390265A CN201910543699.9A CN201910543699A CN110390265A CN 110390265 A CN110390265 A CN 110390265A CN 201910543699 A CN201910543699 A CN 201910543699A CN 110390265 A CN110390265 A CN 110390265A
- Authority
- CN
- China
- Prior art keywords
- identification
- content information
- shooting image
- abnormality
- recognition detection
- 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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/176—Urban or other man-made structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Software Systems (AREA)
- Evolutionary Biology (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Human Computer Interaction (AREA)
- Medical Informatics (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to construction site inspection fields, and in particular to a kind of recognition detection method of unmanned plane inspection, unmanned plane fly and shoot automatically according to polling path, obtain shooting image；Identification shooting image simultaneously obtains its content information, alternatively, identification shoots image and obtains its content information and corresponding structure size；In conjunction with preset rules, content information identification or/and structure size identification are carried out；Difference if it exists is confirmed as recognizing the abnormality of the shooting image, and to identify the abnormality of examined in determination or region, the invention further relates to a kind of recognition detection systems of unmanned plane inspection.The present invention judges the abnormality in identification examined in determination or region, improves recognition efficiency, and improve identification order of accuarcy, brings identification difficulty big specifically for the complex environment in construction site, the not high problem of accuracy rate；Further, it by machine learning, realizes more efficient intelligent recognition, improves the feasible operability of identification.
Description
Technical field
The present invention relates to construction site inspection fields, and in particular to a kind of recognition detection method of unmanned plane inspection and is
System.
Background technique
Construction site inspection is to fully understand building-site condition of construction, by the whole of project inspection quality results
Change and examine, finds and solve relevant issues present in construction in time, improve company's project quality, the control of the targets such as quality
Ability.Make project in construction field engineering quality, safe construction inspection work standardization, degreeization, specialized, the every engineering matter of promotion
Amount and construction, safety management level are continuously improved.
Construction site inspection generally comprises Daily Round Check and safety patrol inspection.
About Daily Round Check, 1, the engineering technology portion professional every workday is required daily patrol to construction operation point
Inspection, Daily Round Check look into morning and afternoon respectively twice.2, independently carried out by engineering department's members, visual condition of construction notice project management department and
Supervising engineer participates in, the implementation of construction quality, progress issue and corresponding measure, is irregularly taken out by leader with specific duties to management log
It looks into.3, daily patrolmans all before leaving offices write out management log, indicate same day inspection situation and practice, record engineering conference inspection
Item, project signature design alteration, the implementation of construction quality, progress issue and corresponding measure are looked into, by leader with specific duties to management day
Will is irregularly spot-check.
About safety patrol, 1, engineering technology portion safety patrol after leading inspection weekly, by engineering technology portion personnel, peace
Crew leads group of people, and project management department sole duty security official and supervision office sole duty security official participate in, and carries out the comprehensive security inspection in building site.Safety inspection
It looks into and accomplishes there is record every time, the accident potential found should be accomplished to determine people, timing, set down measures, fund is implemented and rectify and improve comprehensively.2, right
The problem of finding in inspection security official fills in reform advice and project management department manager and management is transferred to sign for, and special messenger is arranged to be responsible for supervising
Unit in charge of construction's rectification.3, builder and security official should check rectification result, can just continue to construct after qualified.For rectification
Duration longer position, signs and issues rectification notification sheet, and time limit is rectified and improved and checked.
The main contents of construction site inspection include: 1, understand site operation quality, safety, the practicable completion feelings of schedule
Condition, data dynamic, find quality, safety construction, progress and material problem present in construction in time.2, to being sent out in inspection
The incorrect construction method of existing mass defect, technique, the problems such as not implementing by standard, propose reform advice.
But inspection process in construction site is complicated, personnel cost is high, meanwhile, in large-scale inspection, it is easy to produce
It judges by accident or fails to judge, also cannot achieve accurate true inspection in real time, with very big time delay and error.
Intelligent identification technology can also be used, apply in the inspection in construction site, realize intelligent recognition, but due to building
Structure is complicated in building site, always mixes there are many structure, and identification difficulty is big, and recognition accuracy is not high.
Summary of the invention
The technical problem to be solved in the present invention is that in view of the above drawbacks of the prior art, providing a kind of unmanned plane inspection
Recognition detection method and system, the problems such as identification difficulty for solving existing intelligent recognition is big, recognition accuracy is not high.
The technical solution adopted by the present invention to solve the technical problems is: providing a kind of recognition detection side of unmanned plane inspection
The step of method, the preset rules including examined in determination or region, the recognition detection method includes: unmanned plane according to polling path
Automatic flight and shooting obtain shooting image；Identification shooting image simultaneously obtains its content information, alternatively, identification shooting image is simultaneously
Obtain its content information and corresponding structure size；In conjunction with preset rules, carries out content information identification or/and structure size is known
Not；Difference if it exists is confirmed as recognizing the abnormality of the shooting image, to identify the abnormal shape of examined in determination or region
State.
Wherein, preferred version is to be preset with the first identification model based on the content information, the content information identification
The step of include: that content information comparison is carried out according to the first identification model, realize image recognition；Obtain the shooting image
Abnormality, using the abnormality as examined in determination or region.
Wherein, preferred version is that described the step of establishing the first identification model includes: the content of the multiple shooting images of acquisition
Information is data sample；The first identification model based on the content information is established by machine learning in conjunction with preset rules.
Wherein, preferred version is: being preset with the second identification model based on the structure size, the structure size identification
The step of include: that structure size comparison is carried out according to the second identification model, obtain scale error；Obtain the shooting image
Abnormality, using the abnormality as examined in determination or region.
Wherein, preferred version is that described the step of establishing the second identification model includes: the content of the multiple shooting images of acquisition
Information and and corresponding structure size be data sample；It establishes by machine learning in conjunction with preset rules and is based on the structure
Second identification model of size.
Wherein, preferred version be in, the generation step of the polling path include: be preset with based on the examined in determination or
The BIM model in region generates polling path in conjunction with BIM model；Alternatively, the memory manually-operated navigation route of unmanned plane, according to
Presetting navigation route generates polling path.
Wherein, preferred version is: the region of patrolling and examining includes construction site；The examined in determination includes damaged condition, people
Member position, personnel safety cap wear condition, personnel safety are with wear condition, personnel's unlawful practice, the protection feelings for needing protective position
During condition, the normal conditions of design size, the material placement situation of eminence position, the warning protective situation of danger source, violation are built
It is one or more.
Wherein, preferred version is: the step of the recognition detection method further include: one error threshold of setting；In conjunction with default
Rule carries out content information identification or/and structure size identification；Difference if it exists, and difference is more than the range of error threshold, really
Think to recognize the abnormality for shooting image, to identify the abnormality of examined in determination or region.
The technical solution adopted by the present invention to solve the technical problems is: providing a kind of recognition detection system of unmanned plane inspection
System, the recognition detection system includes: processing unit, and the processing unit is stored with computer program, the computer program
The step of capable of being performed to realize the method；Unmanned plane flies according to polling path automatically under control of the processing unit
And shooting, obtain shooting image.
Wherein, preferred version is: the processing unit includes identification shooting image module, recognition processing module and judges mould
Block, the identification shooting image module identification shooting image simultaneously obtains its content information, alternatively, the identification shoots image module
Identification shooting image simultaneously obtains its content information and corresponding structure size；The recognition processing module combination preset rules,
Carry out content information identification or/and structure size identification；The judgment module judgement has differences, and is confirmed as recognizing the bat
The abnormality of image is taken the photograph, to identify the abnormality of examined in determination or region.
The beneficial effects of the present invention are compared with prior art, the present invention is by designing the knowledge of unmanned plane inspection a kind of
Other detection method and system shoot polling path by unmanned plane inspection, in conjunction with preset rules and to shooting image
Content information or dimension information identified, the difference of shooting image is obtained, to judge to identify the different of examined in determination or region
Normal state improves recognition efficiency, and improves identification order of accuarcy, brings identification specifically for the complex environment in construction site
Difficulty is big, the not high problem of accuracy rate；Further, it by machine learning, realizes more efficient intelligent recognition, improves identification
Feasible operability.
Detailed description of the invention
Present invention will be further explained below with reference to the attached drawings and examples, in attached drawing:
Fig. 1 is the flow diagram of recognition detection method of the present invention；
Fig. 2 is the flow diagram of the content of present invention information identification；
Fig. 3 is the flow diagram of structure of the invention Dimensions recognition；
Fig. 4 is the flow diagram that the present invention generates polling path embodiment one；
Fig. 5 is the flow diagram that the present invention generates polling path embodiment two；
Fig. 6 is the flow diagram of the recognition detection method the present invention is based on error threshold；
Fig. 7 is the structural schematic diagram of recognition detection system of the present invention；
Fig. 8 is the structural schematic diagram of Fig. 7.
Specific embodiment
Now in conjunction with attached drawing, elaborate to presently preferred embodiments of the present invention.
As shown in Figure 1, the present invention provides the preferred embodiment of recognition detection method.
A kind of the step of recognition detection method of unmanned plane inspection, the recognition detection method includes:
Step S11, the preset rules including examined in determination or region；
Step S12, unmanned plane flies and shoots automatically according to polling path, obtains shooting image；
Step S131, identification shoots image and obtains its content information；
Step S132, in conjunction with preset rules, content information identification is carried out；
Alternatively, step S141, identification shoot image and obtain its content information and corresponding structure size；
Step S142, in conjunction with preset rules, structure size identification is carried out；
Step S15, difference if it exists is confirmed as recognizing the abnormality of the shooting image, to identify examined in determination
Or the abnormality in region.
Wherein, after completing step S12, may be selected to enter step S131, step S132 and step S15, or select into
Enter step S141, step S142 and step S15, specific choice is determined according to the identification content of shooting image, such as identification shooting
The structure content of image, selects the former, such as the size content of identification shooting image, selects the latter.
Wherein, the region of patrolling and examining includes construction site；The examined in determination includes damaged condition, personnel positions, personnel
Safety cap wear condition, personnel safety are with wear condition, personnel's unlawful practice, protection situation, the design size for needing protective position
Normal conditions, eminence position material place situation, the warning protective situation of danger source, in violation of rules and regulations building one of or it is more
Kind.
Specifically:
About step S11, the preset rules are construction site rules that meet national standards, can be by national government correlation
Agencies dictate setting, or by each company and relevant departments' regulation setting and relevant employee management, behavioral standard, specifically need
It according to different scenes and then to limit, to meet the needs of different scenes.For example, safety net cannot have damaged, employee work
Period want safe wearing cap, high altitude operation need wear safety belt, hole need protective fence and warning sign, face side position need it is anti-
Guardrail, foundation pit need to protect, the connector of outer wall stand needs to be staggered 500mm, high-altitude edge cannot prevent article with preventing dropped material, member
Work carries out smoking during working and quarrel and fight noisily.
About step S12, unmanned plane should include that the Flight main body of unmanned plane, the sensing of RTK module and profession and shooting are set
Standby, RTK module is real-time processing two using RTK (Real-time kinematic, in real time dynamically) carrier phase difference technology
The carrier phase that base station acquires is issued receiver user, is asked by the difference method of a measuring station carrier phase observed quantity
Difference resolves coordinate, this is a kind of new common satellite positioning surveys method, and the inspection road of Flight main body is controlled by RTK module
Diameter.And sensing and capture apparatus include visible light shooting equipment, thermal infrared capture apparatus and laser radar, pass through visible light
Capture apparatus, thermal infrared capture apparatus realize the acquisition of shooting image, by laser radar to prevent collision obstacle.
Control unmanned plane realization barrier avoiding function, hovering function focusing shooting function, flight, breakpoint continue and fly, return automatically manually
Boat, oblique photograph, along the function of conducting wire automatic cruising and ranging.
About step S131 and step S132, identification shooting image simultaneously obtains its content information, i.e., institute is right in shooting image
Structure, position, shape for answering etc., if meet preset rules, carry out content information identification.
Similar to above-mentioned steps S131 and step S132 about step S141 and step S142, only identification shoots image
And its content information and corresponding structure size are obtained, the bases such as corresponding structure, position, shape in the shooting image
On plinth, size comparison is also carried out, whether size meets preset rules, carries out structure size identification.
About step S15, in above-mentioned identification, the difference is construed as being unsatisfactory for the requirement of preset rules, such as pacifies
The whole network cannot have breakage, when shooting image taking safety net, obtain the integral layout of safety net, and identify the whole of safety net
Body effect, if it exists hole or tearing port, then it is assumed that be unsatisfactory for the requirement of preset rules, be considered as having differences at this time.
As shown in Fig. 2, the present invention provides the preferred embodiment of content information identification.
It is preset with the first identification model based on the content information, can refer to step S21 and step S22.
The content information identify the step of include:
Step S1321, according to the first identification model, content information comparison is carried out, realizes image recognition；
Step S15, the abnormality for obtaining the shooting image, using the abnormality as examined in determination or region.
In the present embodiment, described the step of establishing the first identification model, includes:
Step S21, the content information for acquiring multiple shooting images is data sample；
Step S22, the first identification model based on the content information is established by machine learning in conjunction with preset rules.
Meanwhile the preset rules of step S11 are given in step S22, the content information of step S131 gives step S1321,
And step step S1321 is regarded as the specific extension content of step S132.
Specifically:
About step S21 and step S22, machine learning (Machine Learning, ML) is that a multi-field intersection is learned
Section is related to the multiple subjects such as probability theory, statistics, Approximation Theory, convextiry analysis, algorithm complexity theory.Why specialize in computer
Original mold is quasi- or realizes that the learning behavior of the mankind reorganizes the existing structure of knowledge and be allowed to not to obtain new knowledge or skills
The disconnected performance for improving itself.It is the core of artificial intelligence, is the fundamental way for making computer have intelligence, and application spreads people
The every field of work intelligence, it is mainly using conclusion, comprehensive rather than deduction.
Multiple data samples continue to optimize the first identification model by machine learning, improve the identification of the first identification model
Accuracy.
And about step S1321 and step S15, identification to content information, it is most important be first will be in shooting image
Content information extracts, then the content information of extraction is carried out image recognition according to the first identification model, compares otherness, and root
Judge whether to meet preset rules according to otherness, be considered as normally if meeting, it is counter to think that there are abnormalities.
As shown in figure 3, the present invention provides the preferred embodiment of structure size identification.
It is preset with the second identification model based on the structure size, can refer to step S31 and step S32.
The structure size identify the step of include:
Step S1421, according to the second identification model, structure size comparison is carried out, obtains scale error；
Step S15, the abnormality for obtaining the shooting image, using the abnormality as examined in determination or region.
Described the step of establishing the second identification model includes:
Step S31, acquire it is multiple shooting images content informations and and corresponding structure size be data sample；
Step S31, the second identification model based on the structure size is established by machine learning in conjunction with preset rules.
Meanwhile the preset rules of step S11 are given in step S32, the content information of step S141 gives step S1421,
And step step S1421 is regarded as the specific extension content of step S142.
Specifically:
About step S31 and step S32, description is almost the same with above-mentioned steps S21 and step S22, herein no longer one by one
Description.The difference is that: multiple data samples continue to optimize the second identification model by machine learning, improve the second identification
The recognition accuracy of model, and the data sample is the content information and and corresponding structure size of multiple shooting images.
The problem of structure size is not only about image shape structure further includes that dimension scale between different structure is asked
Topic.
And about step S1421 and step S15, identification to structure size, it is most important be first will be in shooting image
Content information extracts, and in the structure size for obtaining the content information, then the structure size that will acquire is according to the second identification model
Image recognition is carried out, compares otherness, and judge whether to meet preset rules according to otherness, is considered as normally if meeting,
It is anti-then think that there are abnormalities.
As shown in Figure 4 and Figure 5, the present invention provides the preferred embodiment for generating polling path.
The generation step of the polling path includes:
Step S41, it is preset with the BIM model based on the examined in determination or region；
Step S42, in conjunction with BIM model, polling path is generated.
Alternatively, step S51, the memory manually-operated navigation route of unmanned plane；
Step S52, polling path is generated according to presetting navigation route.
Generating polling path may be present two approach, be respectively step S41 and step S42 and step step S51 and
Step S52.
Wherein, BIM model, core are by establishing virtual architectural engineering threedimensional model, are this using digitizing technique
A model provides the complete and consistent architectural engineering information bank of actual conditions.The information bank not only includes description building structure
Geological information, professional attributes and the status information of part further comprise the state letter of non-component object (such as space, motor behavior)
Breath.The BIM model based on the examined in determination or region is established by related software, to reflect the examined in determination or region
Three-dimensional structure.
Specifically:
Firstly, selected object or the region for needing inspection, as examined in determination or region, and is directed to examined in determination or region
The core or other critical positions of wanted inspection investigation produce polling path, using the flight path as unmanned plane.
Secondly, the polling path is that multiple location points are arranged on BIM model in step S41 and step S42, it will
Location point is together in series, and eliminates obstruction and calculate best (time is most short or can inspect a little at most), forms a polling path；Weight
Point is in the environment with barrier, according to certain evaluation criterion, finds one from initial state to dbjective state
Collisionless path.
And in step S51 and step S52, after manual flight's operation, remembers as navigation route, can also will navigate by water
Route and BIM model cooperate, and keep the accuracy of navigation route higher.
As shown in fig. 6, the present invention provides the preferred embodiment of the recognition detection method based on error threshold.
The step of recognition detection method further include:
Step S151, an error threshold is set；Wherein, in conjunction with preset rules, content information identification or/and structure ruler are carried out
Very little identification；
Step S152, difference if it exists, and difference is more than the range of error threshold, is confirmed as recognizing the shooting image
Abnormality, to identify the abnormality of examined in determination or region.
Wherein, about carry out content information identification or/and structure size identification specific steps it is above-mentioned specifically
It is bright, it no longer describes one by one herein.
Meanwhile about step S151, be eliminate the error as caused by irresistible factor, keep recognition result more real.
Such as error threshold can be the tolerance of the shape of structure, due to shooting angle and aperture problem, may cause shooting figure
Content information as in generates distortion or deformation, after content information identification, to eliminate brought by the distortion or deformation accidentally
Difference.
Or be light or shooting angle problem, lead to shoot content information in image and corresponding
Deformation occurs for structure size, and ratio generates exception, is at this moment also required to error correction.
Specific error threshold needs to be configured according to practical application scene, can be judged according to the scene of technical staff,
Or the extent of error of machine learning is configured, can also at times, point place, classifying type be configured error threshold；
Error threshold can be a kind of error value range, be also possible to a kind of percentage error range.
As shown in Figure 7 and Figure 8, the present invention provides the preferred embodiment of recognition detection system.
A kind of recognition detection system of unmanned plane inspection, the recognition detection system include processing unit 100 and unmanned plane
200, the processing unit 100 is stored with computer program, and the computer program can be performed to realize the method
Step；It under the control of processing unit 100, flies and shoots automatically according to polling path, obtain shooting image.
Specifically, processing unit 100 includes the preset rules in examined in determination or region, and unmanned plane 200 is in processing unit 100
It under control, flies and shoots automatically according to polling path, obtain shooting image, and be sent in processing unit 100；Processing unit
100 identification shooting images simultaneously obtain its content information, alternatively, the identification of processing unit 100 shoots image and obtains its content information
And corresponding structure size；Processing unit 100 combines preset rules, carries out content information identification or/and structure size identification；
Processing unit 100 judges difference, if it exists difference, is confirmed as recognizing the abnormality of the shooting image, to identify inspection
The abnormality in object or region.
Further, the processing unit 100 includes identification shooting image module 101, recognition processing module 102 and judgement
Module 103, the identification shooting identification of the image module 101 shooting image simultaneously obtains its content information, alternatively, the identification is shot
The identification of image module 101 shooting image simultaneously obtains its content information and corresponding structure size；The recognition processing module 102
In conjunction with preset rules, content information identification or/and structure size identification are carried out；The judgement of judgment module 103 has differences, really
Think to recognize the abnormality for shooting image, to identify the abnormality of examined in determination or region.
Unmanned plane 200 includes unmanned plane main control module 201 and camera 202, and unmanned plane main control module 201 controls unmanned plane
200 fly automatically according to polling path, and control camera 202 is shot, and obtains shooting image.
Wherein, unmanned plane main control module 201 can be set up directly on unmanned plane 200, be also possible to the place for having unmanned plane 200
It manages circuit and remote management platform combination is constituted.
As described above, only preferred embodiment is not intended to limit the scope of the present invention, Fan Yibenfa
Equivalent change or modification made by bright claim is all that the present invention is covered.
Claims (10)
1. a kind of recognition detection method of unmanned plane inspection, which is characterized in that the preset rules including examined in determination or region, institute
The step of stating recognition detection method include:
Unmanned plane flies and shoots automatically according to polling path, obtains shooting image；
Identification shooting image simultaneously obtains its content information, alternatively, identification shooting image and obtaining its content information and corresponding
Structure size；
In conjunction with preset rules, content information identification or/and structure size identification are carried out；
Difference if it exists is confirmed as recognizing the abnormality of the shooting image, to identify the exception of examined in determination or region
State.
2. recognition detection method according to claim 1, which is characterized in that be preset with first based on the content information
Identification model, the content information identify the step of include:
According to the first identification model, content information comparison is carried out, realizes image recognition；
The abnormality for obtaining the shooting image, using the abnormality as examined in determination or region.
3. recognition detection method according to claim 2, which is characterized in that the step of establishing the first identification model packet
It includes:
The content information for acquiring multiple shooting images is data sample；
The first identification model based on the content information is established by machine learning in conjunction with preset rules.
4. recognition detection method according to claim 1, it is characterised in that: be preset with second based on the structure size
Identification model, the structure size identify the step of include:
According to the second identification model, structure size comparison is carried out, obtains scale error；
The abnormality for obtaining the shooting image, using the abnormality as examined in determination or region.
5. recognition detection method according to claim 2, which is characterized in that the step of establishing the second identification model packet
It includes:
Acquire it is multiple shooting images content informations and and corresponding structure size be data sample；
The second identification model based on the structure size is established by machine learning in conjunction with preset rules.
6. recognition detection method according to claim 1, which is characterized in that the generation step of the polling path includes:
It is preset with the BIM model based on the examined in determination or region, in conjunction with BIM model, generates polling path；
Alternatively, the memory manually-operated navigation route of unmanned plane, generates polling path according to presetting navigation route.
7. recognition detection method according to claim 1, it is characterised in that: the region of patrolling and examining includes construction site；Institute
State examined in determination include damaged condition, personnel positions, personnel safety cap wear condition, personnel safety band wear condition, personnel disobey
Rule behavior, the protection situation for needing protective position, the normal conditions of design size, material the placement situation, danger source of eminence position
Warning protective situation, in violation of rules and regulations building one of or it is a variety of.
8. recognition detection method according to claim 1, it is characterised in that: the step of recognition detection method also wraps
It includes:
One error threshold is set；
In conjunction with preset rules, content information identification or/and structure size identification are carried out；
Difference if it exists, and difference is more than the range of error threshold, is confirmed as recognizing the abnormality of the shooting image, with
Identify the abnormality in examined in determination or region.
9. a kind of recognition detection system of unmanned plane inspection, which is characterized in that the recognition detection system includes:
Processing unit, the processing unit are stored with computer program, and the computer program can be performed to realize as weighed
Benefit requires the step of any one of 1 to 8 the method；
Unmanned plane flies and shoots automatically according to polling path under control of the processing unit, obtains shooting image.
10. recognition detection system according to claim 9, it is characterised in that: the processing unit includes identification shooting figure
As module, recognition processing module and judgment module, the identification shooting image module identification shooting image simultaneously obtains its content letter
Breath, alternatively, identification shooting image module identification shoots image and obtains its content information and corresponding structure size；Institute
Recognition processing module combination preset rules are stated, content information identification or/and structure size identification are carried out；The judgment module judgement
It has differences, is confirmed as recognizing the abnormality of the shooting image, to identify the abnormality of examined in determination or region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910543699.9A CN110390265A (en) | 2019-06-21 | 2019-06-21 | A kind of recognition detection method and system of unmanned plane inspection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910543699.9A CN110390265A (en) | 2019-06-21 | 2019-06-21 | A kind of recognition detection method and system of unmanned plane inspection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110390265A true CN110390265A (en) | 2019-10-29 |
Family
ID=68285746
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910543699.9A Pending CN110390265A (en) | 2019-06-21 | 2019-06-21 | A kind of recognition detection method and system of unmanned plane inspection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110390265A (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111726576A (en) * | 2020-05-27 | 2020-09-29 | 深圳中琛源科技股份有限公司 | Unmanned aerial vehicle inspection method, device, system and storage medium |
| CN111754451A (en) * | 2019-12-31 | 2020-10-09 | 广州极飞科技有限公司 | Surveying and mapping unmanned aerial vehicle achievement detection method and device, electronic equipment and storage medium |
| CN111783718A (en) * | 2020-07-10 | 2020-10-16 | 浙江大华技术股份有限公司 | Target object state identification method and device, storage medium and electronic device |
| CN111924119A (en) * | 2020-07-15 | 2020-11-13 | 招商局国际信息技术有限公司 | Wharf intelligent point inspection method and system based on unmanned aerial vehicle |
| CN111929712A (en) * | 2020-05-20 | 2020-11-13 | 国网浙江省电力有限公司 | Engineering site safety monitoring method based on RTK technology |
| CN112307913A (en) * | 2020-10-20 | 2021-02-02 | 江苏濠汉信息技术有限公司 | Protective equipment wearing detection method and device based on unmanned aerial vehicle vision |
| CN113095141A (en) * | 2021-03-15 | 2021-07-09 | 南通大学 | Unmanned aerial vehicle vision learning system based on artificial intelligence |
| CN113204247A (en) * | 2021-04-16 | 2021-08-03 | 深圳市艾赛克科技有限公司 | Unmanned aerial vehicle system of patrolling and examining |
| CN113433971A (en) * | 2021-07-09 | 2021-09-24 | 深圳大学 | Method, device, equipment and storage medium for acquiring data of high-rise building exterior wall |
| CN113762183A (en) * | 2021-09-13 | 2021-12-07 | 墙管家建筑科技(上海)有限公司 | Intelligent checking and analyzing system for existing building safety and operation method |
| CN113837115A (en) * | 2021-09-28 | 2021-12-24 | 上海翼枭航空科技有限公司 | Roof solar energy identification method and system based on unmanned aerial vehicle |
| CN114119567A (en) * | 2021-11-30 | 2022-03-01 | 深圳大学 | Human-computer interaction type intelligent detection method for outer wall diseases of high-rise building |
| CN114549984A (en) * | 2022-02-22 | 2022-05-27 | 深圳市红湾安全智能科技有限公司 | Intelligent management method and system for battery car |
| CN114758260A (en) * | 2022-06-15 | 2022-07-15 | 成都鹏业软件股份有限公司 | Construction site safety protection net detection method and system |
| CN114967760A (en) * | 2022-07-20 | 2022-08-30 | 无锡建设监理咨询有限公司 | Building engineering supervision method and system based on unmanned aerial vehicle and storage medium |
| CN116301055A (en) * | 2023-04-25 | 2023-06-23 | 西安玖安科技有限公司 | Unmanned aerial vehicle inspection method and system based on building construction |
| CN116389695A (en) * | 2023-06-06 | 2023-07-04 | 深圳市城市公共安全技术研究院有限公司 | Building site monitoring method and device, building site inspection equipment and storage medium |
| CN117312591A (en) * | 2023-10-17 | 2023-12-29 | 南京海汇装备科技有限公司 | Image data storage management system and method based on virtual reality |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106379535A (en) * | 2016-09-29 | 2017-02-08 | 深圳大学 | Modern tram overhead contact system inspecting method and device based on unmanned aerial vehicle |
| CN106791597A (en) * | 2016-11-11 | 2017-05-31 | 韦红兰 | A kind of unmanned plane that crop can be monitored |
| CN109557935A (en) * | 2018-11-01 | 2019-04-02 | 广州圣安环保科技有限公司 | A kind of safety monitoring during construction method and system based on unmanned plane |
| CN109840600A (en) * | 2018-12-29 | 2019-06-04 | 天津大学 | The feeder channel unmanned plane of BIM auxiliary cooperates with cruising inspection system online |
-
2019
- 2019-06-21 CN CN201910543699.9A patent/CN110390265A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106379535A (en) * | 2016-09-29 | 2017-02-08 | 深圳大学 | Modern tram overhead contact system inspecting method and device based on unmanned aerial vehicle |
| CN106791597A (en) * | 2016-11-11 | 2017-05-31 | 韦红兰 | A kind of unmanned plane that crop can be monitored |
| CN109557935A (en) * | 2018-11-01 | 2019-04-02 | 广州圣安环保科技有限公司 | A kind of safety monitoring during construction method and system based on unmanned plane |
| CN109840600A (en) * | 2018-12-29 | 2019-06-04 | 天津大学 | The feeder channel unmanned plane of BIM auxiliary cooperates with cruising inspection system online |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111754451A (en) * | 2019-12-31 | 2020-10-09 | 广州极飞科技有限公司 | Surveying and mapping unmanned aerial vehicle achievement detection method and device, electronic equipment and storage medium |
| CN111929712A (en) * | 2020-05-20 | 2020-11-13 | 国网浙江省电力有限公司 | Engineering site safety monitoring method based on RTK technology |
| CN111726576A (en) * | 2020-05-27 | 2020-09-29 | 深圳中琛源科技股份有限公司 | Unmanned aerial vehicle inspection method, device, system and storage medium |
| CN111783718A (en) * | 2020-07-10 | 2020-10-16 | 浙江大华技术股份有限公司 | Target object state identification method and device, storage medium and electronic device |
| CN111924119A (en) * | 2020-07-15 | 2020-11-13 | 招商局国际信息技术有限公司 | Wharf intelligent point inspection method and system based on unmanned aerial vehicle |
| CN112307913A (en) * | 2020-10-20 | 2021-02-02 | 江苏濠汉信息技术有限公司 | Protective equipment wearing detection method and device based on unmanned aerial vehicle vision |
| CN112307913B (en) * | 2020-10-20 | 2021-09-28 | 江苏濠汉信息技术有限公司 | Protective equipment wearing detection method and device based on unmanned aerial vehicle vision |
| CN113095141A (en) * | 2021-03-15 | 2021-07-09 | 南通大学 | Unmanned aerial vehicle vision learning system based on artificial intelligence |
| CN113204247A (en) * | 2021-04-16 | 2021-08-03 | 深圳市艾赛克科技有限公司 | Unmanned aerial vehicle system of patrolling and examining |
| CN113433971A (en) * | 2021-07-09 | 2021-09-24 | 深圳大学 | Method, device, equipment and storage medium for acquiring data of high-rise building exterior wall |
| CN113762183A (en) * | 2021-09-13 | 2021-12-07 | 墙管家建筑科技(上海)有限公司 | Intelligent checking and analyzing system for existing building safety and operation method |
| CN113837115A (en) * | 2021-09-28 | 2021-12-24 | 上海翼枭航空科技有限公司 | Roof solar energy identification method and system based on unmanned aerial vehicle |
| CN114119567A (en) * | 2021-11-30 | 2022-03-01 | 深圳大学 | Human-computer interaction type intelligent detection method for outer wall diseases of high-rise building |
| CN114549984A (en) * | 2022-02-22 | 2022-05-27 | 深圳市红湾安全智能科技有限公司 | Intelligent management method and system for battery car |
| CN114758260A (en) * | 2022-06-15 | 2022-07-15 | 成都鹏业软件股份有限公司 | Construction site safety protection net detection method and system |
| CN114758260B (en) * | 2022-06-15 | 2022-10-18 | 成都鹏业软件股份有限公司 | Construction site safety protection net detection method and system |
| CN114967760A (en) * | 2022-07-20 | 2022-08-30 | 无锡建设监理咨询有限公司 | Building engineering supervision method and system based on unmanned aerial vehicle and storage medium |
| CN114967760B (en) * | 2022-07-20 | 2024-01-16 | 无锡建设监理咨询有限公司 | Unmanned plane-based building engineering supervision method, system and storage medium |
| CN116301055A (en) * | 2023-04-25 | 2023-06-23 | 西安玖安科技有限公司 | Unmanned aerial vehicle inspection method and system based on building construction |
| CN116389695A (en) * | 2023-06-06 | 2023-07-04 | 深圳市城市公共安全技术研究院有限公司 | Building site monitoring method and device, building site inspection equipment and storage medium |
| CN116389695B (en) * | 2023-06-06 | 2023-09-01 | 深圳市城市公共安全技术研究院有限公司 | Building site monitoring method and device, building site inspection equipment and storage medium |
| CN117312591A (en) * | 2023-10-17 | 2023-12-29 | 南京海汇装备科技有限公司 | Image data storage management system and method based on virtual reality |
| CN117312591B (en) * | 2023-10-17 | 2024-03-12 | 南京海汇装备科技有限公司 | Image data storage management system and method based on virtual reality |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110390265A (en) | A kind of recognition detection method and system of unmanned plane inspection | |
| CN110398982A (en) | A kind of method for inspecting and system of unmanned plane | |
| CN206672041U (en) | A kind of mining equipment intelligent patrol detection based on augmented reality and safeguard accessory system | |
| CN108319926A (en) | A kind of the safety cap wearing detecting system and detection method of building-site | |
| CN101751744B (en) | Detection and early warning method of smoke | |
| CN111310947A (en) | Building facility operation and maintenance method, equipment, storage medium and system based on 5G | |
| CN108710126A (en) | Automation detection expulsion goal approach and its system | |
| CN113903081A (en) | Visual identification artificial intelligence alarm method and device for images of hydraulic power plant | |
| CN108389359B (en) | Deep learning-based urban fire alarm method | |
| CN114155601A (en) | Vision-based method and system for detecting dangerous behaviors of operating personnel | |
| CN113807240A (en) | Intelligent transformer substation personnel dressing monitoring method based on uncooperative face recognition | |
| CN114863489B (en) | Virtual reality-based movable intelligent auxiliary inspection method and system for construction site | |
| CN112235537A (en) | Transformer substation field operation safety early warning method | |
| CN111950359A (en) | Intelligent system and method for preventing human errors of nuclear power plant | |
| CN111753780A (en) | Transformer substation violation detection system and violation detection method | |
| CN113256269A (en) | Engineering management system based on BIM, cloud computing and big data technology | |
| CN114897507A (en) | Building construction management system and management method based on BIM | |
| CN113713292A (en) | Method and device for carrying out accurate flame discrimination, fire extinguishing point positioning and rapid fire extinguishing based on YOLOv5 model | |
| CN112949457A (en) | Maintenance method, device and system based on augmented reality technology | |
| CN111091104A (en) | Target object protection detection method, device, equipment and storage medium | |
| CN115169777A (en) | Intelligent supervision system for safety operation | |
| Rivera et al. | Work safety assessment through contextual analysis with computer vision | |
| CN116665419B (en) | Intelligent fault early warning system and method based on AI analysis in power production operation | |
| CN112861737A (en) | Forest fire smoke detection method based on image dark channel and YoLov3 | |
| CN115082765B (en) | On-site operation management method and on-site operation management system |
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: 20191029 |