CN113537159B - Crane risk data identification method based on artificial intelligence - Google Patents
Crane risk data identification method based on artificial intelligence Download PDFInfo
- Publication number
- CN113537159B CN113537159B CN202111057534.4A CN202111057534A CN113537159B CN 113537159 B CN113537159 B CN 113537159B CN 202111057534 A CN202111057534 A CN 202111057534A CN 113537159 B CN113537159 B CN 113537159B
- Authority
- CN
- China
- Prior art keywords
- cotter
- risk
- cotter pin
- characteristic value
- nut groove
- 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.)
- Active
Links
Images
Abstract
Hair brushThe crane risk data identification method based on artificial intelligence comprises the following steps: step (ii) of
: automatically acquiring a real-time image of a cotter area of the tower crane by using an unmanned aerial vehicle, and identifying and dividing a device where the cotter is located; step (ii) of
: continuously analyzing and identifying the cotter pin and the nut groove according to the segmentation image to obtain the image characteristics of each part; step (ii) of
: constructing an evaluation index characteristic value reflecting the cotter pin risk, and evaluating the cotter pin risk degree of the current crane; step (ii) of
: whether the obtained cotter risk degree evaluation value is larger than a threshold value or not is detected, whether early warning operation is conducted or not is judged, and therefore dangerous accidents are avoided. The method and the device realize self-adaptive detection and intelligent evaluation of the safety risk of the cotter pin of the tower crane in the high-altitude environment, and can effectively prevent dangerous accidents by early warning.
Description
Technical Field
The application relates to the technical field of risk data identification, in particular to a crane risk data identification method based on artificial intelligence.
Background
The crane has more potential safety hazards in the use process, so the crane needs to be checked frequently, but once the equipment is built, the crane generally needs to operate in an overhead environment for a long time, so the manual detection is difficult to realize. Particularly, if the small component is not opened or the opening degree is insufficient, the small component can easily fall off in the using process, so that the pin shaft without the split pin automatically falls off in use, and finally a major accident of arm folding can be caused, and therefore, the crane risk data identification through artificial intelligence is very necessary.
Disclosure of Invention
Aiming at the problems, the invention provides a crane risk data identification method based on artificial intelligence, which comprises the following steps:
step (ii) of
: automatically acquiring a real-time image of a cotter area of the tower crane by using an unmanned aerial vehicle, and identifying and dividing a device where the cotter is located;
step (ii) of
: continuously analyzing and identifying the cotter pin and the nut groove according to the segmentation image to obtain the image characteristics of each part;
step (ii) of
: constructing an evaluation index characteristic value reflecting the cotter pin risk, and evaluating the cotter pin risk degree of the current crane;
step (ii) of
: whether the obtained cotter risk degree evaluation value is larger than a threshold value or not is detected, whether early warning operation is conducted or not is judged, and therefore dangerous accidents are avoided.
Has the advantages that:
(1) according to the invention, the crane risk data is identified through artificial intelligence, the self-adaptive detection and intelligent evaluation of the safety risk of the cotter pin of the tower crane in a high-altitude environment are realized, and the occurrence of dangerous accidents can be effectively prevented through early warning.
(2) The invention is based on artificial intelligence, but does not need to use a neural network, can realize the detection of the working state of the cotter pin, reduces the workload, and thus improves the detection efficiency.
Drawings
Fig. 1 is a schematic diagram of an image of a target acquired by an unmanned aerial vehicle in the crane risk data identification method based on artificial intelligence provided by the invention.
Detailed Description
In order to make the present invention more comprehensible to those skilled in the art, the present invention is described below with reference to examples and the accompanying drawings.
In order to realize the content, the invention designs a crane risk data identification method based on artificial intelligence, which comprises the following steps:
step (ii) of
: and automatically acquiring a real-time image of a cotter area of the tower crane by using the unmanned aerial vehicle, and identifying and segmenting the device where the cotter is located.
Because the operation environment of the tower crane is high altitude, the unmanned aerial vehicle is used for collecting the target image by using the RGB camera. Meanwhile, in order to save human resources, the unmanned aerial vehicle is used for acquiring the target image in a self-adaptive mode.
The method comprises the steps of firstly obtaining a three-dimensional model of the tower crane to be detected according to prior data, then obtaining an imaging visual angle of a cotter pin area through the three-dimensional model, and matching in the three-dimensional model by utilizing real-time imaging of the unmanned aerial vehicle. The finally obtained target image is shown in fig. 1, wherein the matching process includes performing edge detection on the target image, and performing normalized similarity matching according to the obtained edge contour and the contour under the calibrated view angle in the three-dimensional model. When the similarity degree of each contour on the tower crane and the most similar contour of the contour on the tower crane reaches more than 0.8, the real-time image of the cotter pin area can be shot well under the current visual angle.
The grey value of the device where the cotter pin is located is obviously different from that of other regions of the tower crane, so that the device region where the cotter pin is located is segmented based on an Ostu threshold segmentation algorithm, the Ostu threshold segmentation algorithm is an efficient algorithm for carrying out binarization on an image, and is also an adaptive threshold determination method, namely an Otsu threshold segmentation method, which is optimal segmentation in the least square sense. In the pixels larger than the gray segmentation threshold, the region with the largest connected domain area is the device region where the cotter pin is located, otherwise, the region is the background region. Thus, the division and identification of the device area where the cotter pin is located can be realized.
Step (ii) of
: and continuously analyzing and identifying the cotter pin and the nut groove according to the segmentation image to obtain the image characteristics of each part.
Since the device area where the cotter pin is located includes not only the cotter pin but also the nut groove, the subsequent analysis and judgment need to be performed according to different characteristics of the cotter pin and the nut groove, and therefore, the segmented image needs to be continuously analyzed to effectively identify the cotter pin portion and the nut groove portion.
And analyzing a connected domain of the device region where the cotter pin is located obtained through threshold segmentation so as to obtain the external contour information of the region. And counting the number of pixels belonging to the background image in the contour by taking the contour information as a boundary, and selecting a maximum connected region formed by the pixels of the background image in consideration of the influence of noise, wherein the maximum connected region is a closed region formed by the cotter pin at the upper half part of the nut groove.
According to the obtained closed area, boundary lines formed by the area and other areas except the area in the device area where the cotter pin is located can be further obtained, one pixel in the boundary line is arbitrarily selected as a traversal starting point, traversal search is carried out in the contour edge lines of the other areas along a single direction, traversal is finished when the traversed pixel points return to the traversal starting point again, a contour edge contour route with the minimum pixel traversal number is obtained in the process, and the closed area determined by the route is the nut groove area.
The closed area of the cotter pin in the upper half of the nut groove and the area of the nut groove can be obtained, and the remaining part in the area of the device of the cotter pin is the lower half of the cotter pin extending out of the nut groove.
Step (ii) of
: and constructing an evaluation index characteristic value reflecting the cotter pin risk, and evaluating the cotter pin risk degree of the current crane.
The characteristics of each part of the cotter pin are reflected through constructionEvaluation index characteristic value of mouth and sales risk
The risk degree of the cotter pin of the current crane is evaluated, and the evaluation index characteristic value is used for evaluating the risk degree of the cotter pin of the current crane
Are shared by
The three parts are as follows.
The opening angle of the cotter pin is principally the same as when fixing the slotted nut
Therefore, in the invention, the opening angle of the current cotter to be detected is detected firstly, so as to obtain the characteristic value reflecting the risk of the opening angle of the cotter
The specific process is as follows:
in the image of the lower half part of the cotter pin, taking the average value of all pixel coordinates in all pixels of the boundary line of the part and the nut groove area to obtain an initial central pixel point.
And calculating the slope of a straight line formed by the initial central pixel point and all pixel points except the boundary line with the nut groove region in the lower half part image of the cotter pin. All the obtained slope values are processed
Mean value clustering, wherein
Thereby obtaining respective divided portions of the lower half of the cotter pin.
Curvature gradient calculation is carried out on edge pixel points of each part, pixel points at extreme point positions are marked, and in two areas of the lower half part of the cotter pin, only the initial central pixel point and the first extreme point of each area are calculatedThe slope of the formed straight lines is determined, and the included angle formed by the two straight lines at the initial central pixel point is determined according to the slope
And a standard included angle is set to
From this, a characteristic value is obtained which reflects the risk of the opening angle of the cotter pin
The mathematical expression is:
the above formula indicates the included angle
Angle relative to standard
The greater the risk, the lower the risk, and vice versa.
Since the angle can only ensure one aspect of the cotter pin safety, and the symmetry of the cotter pin is also an important point to ensure, the invention establishes the risk characteristic value reflecting the cotter pin opening symmetry according to the aspect of the invention
The specific process is as follows:
calculating the direction of the main component of the cotter pin in the upper half of the nut groove
And principal component directions corresponding to two regions of the cotter pin located at the lower half portion of the nut groove
。
Calculating principal component directions separately
In the direction of principal component
Angle of (2)
Wherein the orientation of the principal component is defined in the present invention
The left values are positive values and the right values are negative values. Under the condition of symmetry
From which a characteristic value can be established
The mathematical expression of (a) is:
the above formula indicates that the more symmetrical the split portion of the split pin is, the less risky, and vice versa.
In addition, the effective length of the lower half part of the split pin is also an important point influencing the risk, so that the risk characteristic value reflecting the effective length of the nut groove is also established in the invention
The specific process is as follows:
according to the main component direction obtained in the above-mentioned process
Calculating its direction to principal component
Is included angle formed by the normal direction of
Then calculating the projection length of the two areas of the lower half part of the cotter pin under the angle
。
So far, the characteristic value can be established according to the obtained projection length
The mathematical expression of (a) is:
the above equation indicates that the longer the effective length of the lower half of the cotter, the lower the risk and vice versa.
It should be noted that the projection lengths calculated in the above process are both positive values, and when two principal component directions are in the same direction
Are all located in the main component direction
When the same side is used, only the larger value is taken as the exponential term in the formula, namely the larger value is taken as the exponential term in the formula
。
So far, the characteristic value obtained in the process can be used
Calculating to obtain characteristic value
The mathematical expression is:
step (ii) of
Whether the obtained cotter risk degree evaluation value is larger than a threshold value or not is detected, whether early warning operation is conducted or not is judged, and therefore dangerous accidents are avoided.
Because the high-altitude operation environment is complex, the rescue difficulty is high, and the damage degree after an accident is large, the threshold value of the cotter risk degree evaluation value is set in the invention
I.e. when the evaluation index characteristic value is detected
When the value of (1) exceeds the threshold value, early warning is required immediately, and workers are informed to carry out subsequent treatment in time, so that the occurrence of harm accidents is prevented.
Claims (6)
1. A crane risk data identification method based on artificial intelligence is characterized by comprising the following steps;
step S1: automatically acquiring a real-time image of a cotter area of the tower crane by using an unmanned aerial vehicle, and identifying and dividing a device where the cotter is located;
step S2: continuously analyzing and identifying the cotter pin and the nut groove according to the segmentation image to obtain the image characteristics of each part;
step S3: when the slotted nut is fixed, the opening angle of the current cotter to be detected is detected according to the image characteristics so as to obtain a characteristic value reflecting the danger of the opening angle of the cotter
(ii) a Characteristic value
The acquisition process comprises the following steps:
1) in the image of the lower half part of the cotter pin, taking the average value of all pixel coordinates in all pixels of the boundary line of the part and the nut groove area to obtain an initial central pixel point;
2) calculating the slope of a straight line formed by the initial central pixel point and all pixel points except the boundary line with the nut groove region in the lower half part image of the cotter pin, and performing the calculation on all obtained slope valueskMean clustering to obtain respective separate portions of the lower half of the cotter;
3) curvature gradient calculation is carried out on edge pixel points of each part, pixel points at extreme point positions are marked, in two areas of the lower half portion of the cotter pin, the slope of a straight line formed by an initial central pixel point and the first extreme point of each area is only calculated, and therefore the included angle formed by the two straight lines at the initial central pixel point is determined
And a standard included angle is set to
Thereby obtaining a characteristic value reflecting the risk of the opening angle of the cotter
The mathematical expression is:
the above formula indicates the included angle
Angle relative to standard
The larger the risk, the lower the risk, otherwise the higher the risk;
establishing a risk characteristic value reflecting the symmetry of the split pin opening
(ii) a Establishing a risky characteristic value reflecting the effective length of the nut groove
(ii) a By a characteristic value
Characteristic value of
And a characteristic value
The average value of the numerical value is used as an evaluation index characteristic value for reflecting the cotter pin risk, and the cotter pin risk degree of the current crane is evaluated;
step S4: whether the obtained cotter risk degree evaluation value is larger than a threshold value or not is detected, whether early warning operation is conducted or not is judged, and therefore dangerous accidents are avoided.
2. The artificial intelligence based crane risk data identification method according to claim 1, wherein the step S1 is as follows:
1) collecting a target image by using an RGB camera by using an unmanned aerial vehicle, and acquiring the target image in a self-adaptive manner;
2) acquiring a three-dimensional model of the tower crane to be detected according to the prior data, then acquiring an imaging visual angle of a cotter pin area through the three-dimensional model, and matching in the three-dimensional model by utilizing real-time imaging of the unmanned aerial vehicle to finally obtain a target image;
according to the characteristic that the gray value of the device where the cotter is located is obviously different from that of other regions of the tower crane, the region of the device where the cotter is located is divided based on an Ostu threshold value division algorithm, in pixels larger than a gray value division threshold value, the region with the largest connected domain area is the region of the device where the cotter is located, and otherwise, the region is a background region, so that division and identification of the region of the device where the cotter is located are achieved.
3. The artificial intelligence based crane risk data identification method according to claim 2, wherein the step S2 is as follows:
1) subsequent analysis and judgment are carried out according to different characteristics of the cotter pin and the nut groove, and the segmentation image is continuously analyzed so as to effectively identify the cotter pin part and the nut groove part;
2) analyzing a connected domain of a device region where the cotter pin is located, wherein the device region is obtained through threshold segmentation, so as to obtain external contour information of the region, counting the number of pixels belonging to a background image in the contour by taking the contour information as a boundary, and selecting a maximum connected domain region formed by the pixels of the background image in consideration of the influence of noise, wherein the maximum connected domain region is a closed region formed by the cotter pin located at the upper half part of the nut groove;
3) according to the obtained closed area, boundary lines formed by the area and other areas except the area in the device area where the cotter pin is located are further obtained, one pixel in the boundary lines is randomly selected to serve as a traversal starting point, traversal search is conducted in the contour edge lines of the other areas along a single direction, traversal is finished when the traversed pixel points return to the traversal starting point again, a contour edge contour route with the minimum pixel traversal number is obtained in the process, and the closed area determined by the route is the nut groove area;
4) the recording obtains the area of the nut groove and the closed area formed by the cotter pin in the upper half part of the nut groove, and the remaining part in the device area of the cotter pin is the lower half part of the cotter pin extending out of the nut groove.
4. The artificial intelligence based crane risk data identification method according to claim 3, wherein the characteristic value reflecting the risk of the opening angle of the cotter is obtained in the step S3
Then, a risk characteristic value reflecting the opening symmetry of the cotter pin is established
The specific process is as follows:
1) calculating the direction of the main component of the cotter pin in the upper half of the nut groovep 1 And principal component directions corresponding to two regions of the cotter pin located at the lower half portion of the nut groovep 2 ,p 3 ；
2) Calculating principal component directions separatelyp 2 ,p 3 Angle with main component direction
Is defined to be located in the main component directionp 1 The left side is positive, the right side is negative, and the symmetry is true
The mathematical expression for the characteristic values is thus established as:
the above formula indicates that the more symmetrical the split portion of the split pin is, the less risky, and vice versa.
5. Artificial intelligence based crane risk data identification method according to claim 4, characterized by the steps of
3 obtaining a characteristic value of the risk reflecting the symmetry of the split pin opening
Then, a risky characteristic value reflecting the effective length of the nut groove is established
The specific process is as follows:
1) according to the main component direction obtained in the above-mentioned processp 2 ,p 3 Calculating its direction to principal componentp 1 Is included angle formed by the normal direction of
Then calculating the projection length of the two areas of the lower half part of the cotter pin under the angle
；
2) The mathematical expression for establishing the characteristic value according to the obtained projection length is as follows:
the above formula shows that the longer the effective length of the lower half part of the cotter, the lower the risk, and vice versa;
wherein, the projection lengths calculated in the above process are all positive values, and when the two principal component directions are in the same directionp 2 ,p 3 Are all located in the main component directionp 1 When the same side is used, only the larger value is taken as the exponential term in the formula, namely the larger value is taken as the exponential term in the formula
。
6. The artificial intelligence based crane risk data identification method according to claim 5, wherein the threshold value of the cotter risk degree assessment value is established in the step S4
Namely, when the value of the evaluation index characteristic value is detected to exceed the threshold value, an early warning is required immediately.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111057534.4A CN113537159B (en) | 2021-09-09 | 2021-09-09 | Crane risk data identification method based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111057534.4A CN113537159B (en) | 2021-09-09 | 2021-09-09 | Crane risk data identification method based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113537159A CN113537159A (en) | 2021-10-22 |
| CN113537159B true CN113537159B (en) | 2021-12-03 |
Family
ID=78093151
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111057534.4A Active CN113537159B (en) | 2021-09-09 | 2021-09-09 | Crane risk data identification method based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113537159B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103559703A (en) * | 2013-10-08 | 2014-02-05 | 中南大学 | Crane barrier monitoring and prewarning method and system based on binocular vision |
| CN104657708A (en) * | 2015-02-02 | 2015-05-27 | 郑州酷派电子设备有限公司 | Novel device and method for identifying three-dimensional object |
| CN106326894A (en) * | 2016-08-31 | 2017-01-11 | 西南交通大学 | Adverse state detection method of transverse pins of rotation double lugs of high-speed rail overhead contact line equipment |
| CN112348034A (en) * | 2020-10-21 | 2021-02-09 | 中电鸿信信息科技有限公司 | Crane defect detection system based on unmanned aerial vehicle image recognition and working method |
-
2021
- 2021-09-09 CN CN202111057534.4A patent/CN113537159B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103559703A (en) * | 2013-10-08 | 2014-02-05 | 中南大学 | Crane barrier monitoring and prewarning method and system based on binocular vision |
| CN104657708A (en) * | 2015-02-02 | 2015-05-27 | 郑州酷派电子设备有限公司 | Novel device and method for identifying three-dimensional object |
| CN106326894A (en) * | 2016-08-31 | 2017-01-11 | 西南交通大学 | Adverse state detection method of transverse pins of rotation double lugs of high-speed rail overhead contact line equipment |
| CN112348034A (en) * | 2020-10-21 | 2021-02-09 | 中电鸿信信息科技有限公司 | Crane defect detection system based on unmanned aerial vehicle image recognition and working method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113537159A (en) | 2021-10-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109522793B (en) | Method for detecting and identifying abnormal behaviors of multiple persons based on machine vision | |
| CN113658132B (en) | Computer vision-based structural part weld joint detection method | |
| CN112434578B (en) | Mask wearing normalization detection method, mask wearing normalization detection device, computer equipment and storage medium | |
| CN112560722B (en) | Airplane target identification method and device, computer equipment and storage medium | |
| US20120213440A1 (en) | Systems and Methods for Automatically Identifying Shadows in Images | |
| CN113370977B (en) | Intelligent vehicle forward collision early warning method and system based on vision | |
| CN112819094A (en) | Target detection and identification method based on structural similarity measurement | |
| KR100752640B1 (en) | Method and apparatus for segmenting fingerprint region using directional gradient filters | |
| CN108563979B (en) | Method for judging rice blast disease conditions based on aerial farmland images | |
| CN111553214B (en) | Method and system for detecting smoking behavior of driver | |
| CN110427979B (en) | Road water pit identification method based on K-Means clustering algorithm | |
| CN117392733B (en) | Acne grading detection method and device, electronic equipment and storage medium | |
| CN104915642A (en) | Method and apparatus for measurement of distance to vehicle ahead | |
| CN111460917B (en) | Airport abnormal behavior detection system and method based on multi-mode information fusion | |
| CN113962994B (en) | Method for detecting cleanliness of lock pin on three-connecting-rod based on image processing | |
| CN113179389A (en) | System and method for identifying crane jib of power transmission line dangerous vehicle | |
| CN113537159B (en) | Crane risk data identification method based on artificial intelligence | |
| CN111582084B (en) | Weak supervision learning-based rail foreign matter detection method and system under empty base view angle | |
| CN117197700A (en) | Intelligent unmanned inspection contact net defect identification system | |
| CN117392611A (en) | Site safety monitoring method, system, equipment and storage medium | |
| CN112734702A (en) | Bridge safety monitoring method based on big data analysis and machine vision and cloud platform | |
| CN111582278B (en) | Portrait segmentation method and device and electronic equipment | |
| CN108960181A (en) | Black smoke vehicle detection method based on multiple dimensioned piecemeal LBP and Hidden Markov Model | |
| JP6851246B2 (en) | Object detector | |
| Cao et al. | Method of Unsupervised Static Recognition and Dynamic Tracking for Vehicles. |
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 |