CN108805207A - A kind of Large Construction vehicle raises arm detection algorithm - Google Patents
A kind of Large Construction vehicle raises arm detection algorithm Download PDFInfo
- Publication number
- CN108805207A CN108805207A CN201810608932.2A CN201810608932A CN108805207A CN 108805207 A CN108805207 A CN 108805207A CN 201810608932 A CN201810608932 A CN 201810608932A CN 108805207 A CN108805207 A CN 108805207A
- Authority
- CN
- China
- Prior art keywords
- large construction
- construction vehicle
- vehicle
- automobile body
- fasterrcnn
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
Abstract
The invention discloses a kind of Large Construction vehicles to raise arm detection algorithm.It includes the following steps：The training that collects pictures detects the FasterRCNN models of Large Construction vehicle；The training that collects pictures detects the FasterRCNN models of Large Construction automobile body；Large Construction vehicle is detected using Faster RCNN algorithms；Large Construction automobile body is detected using Faster RCNN algorithms；The ratio of Large Construction automobile body area and entire vehicle area is calculated, such as less than certain threshold value is then determined as that raising arm is then labeled, and does not otherwise mark.What the present invention can detect Large Construction vehicle raises arm behavior, reaches better Detection accuracy.
Description
Technical field
The invention belongs to transmission line of electricity external force damage prevention field, it is related to a kind of Large Construction vehicle and raises arm detection algorithm.
Background technology
Transmission line of electricity Large Construction vehicle detecting algorithm based on image procossing carries out the intrusion target in passway for transmitting electricity
Identification classification, difficult point essentially consist in target identification, that is, judge whether moving target is the Large Constructions vehicle such as crane, cement pump truck
?.Document【The realization for the intelligent early-warning function that anti-big machinery external force is destroyed in transmission line of electricity】It is middle to be carried out using background subtraction method
Background detection, and characteristics of image of the color as big machinery is used, it is modeled.According to arm region area, bias
Rate and tight ness rating carry out arm detection, using HOUGH transformation calculations arm stretching angles, are sent out according at a distance between transmission line of electricity
Go out alarm.Document【The intelligent measurement of moving target and identification in transmission line of electricity monitoring system】Using the feature based on color preceding
Big machinery is identified in scape target, then positions crane wheel, using 3 area, eccentricity and tight ness rating indexs as classification
The input of device feature vector, to identify arm, and hazard recognition behavior in turn.Method in this two documents all uses color special
Sign carries out target identification, it is believed that the color of Large Construction vehicle is yellow.In practical applications, when there is the large size of non-yellow
When Construction traffic, effect is with regard to very poor.
Invention content
Of the existing technology in order to overcome the problems, such as, the object of the present invention is to provide a kind of Large Construction vehicles to raise arm detection
Algorithm can carry out the identification of Large Construction vehicle and main body car body, and detect Large Construction vehicle raises arm behavior, reaches more preferable
Detection accuracy.
The purpose of the present invention is achieved through the following technical solutions：
A kind of Large Construction vehicle raises arm detection algorithm, it is characterised in that includes the following steps：
(1) training that collects pictures detects the FasterRCNN models of Large Construction vehicle；The Large Construction vehicle packet
Include crane, excavator, cement pump truck etc.；
(2) training that collects pictures detects the FasterRCNN models of Large Construction automobile body；
(3) Faster RCNN algorithms are used to detect Large Construction vehicle；
(4) Faster RCNN algorithms are used to detect Large Construction automobile body；
(5) ratio for calculating Large Construction automobile body area and entire vehicle area, is then judged to raising less than certain threshold value
Arm is then labeled, and is not otherwise marked；The specific steps are：
1) a Large Construction vehicle V identified is chosen respectivelyiThe Large Construction automobile body VB identified with onej
2) judge ViAnd VBjClassification liAnd bljIf they are unequal, step 1) is gone to, is otherwise carried out in next step
3) Large Construction vehicle V is calculatediBand of position piWith Large Construction automobile body VBjBand of position bpjPhase
Hand over region spij(xstart,ystart,xend,yend), computational methods spij(xstart,ystart,xend,yend)=(max (xi start,
xj start),max(yi start,yj start),min(xi end,xj end),min(yi end,yj end)), such as xend<xstartOr yend<ystart, then
Indicate ViBand of position piWith VBjBand of position bpjThere is no intersection, goes to step 1) at this time；
4) vehicle body area including degree c is calculatedij=A (spij)/A(bpj), wherein A (*) indicates the area of band of position *, such as
cij<t1, then it is assumed that Large Construction vehicle V at this timeiWith Large Construction automobile body VBjIncluding degree it is inadequate, be not belonging to same
Vehicle turns to step 1) at this time；
Calculate vehicle body area coincidence factor rij=A (spij)/A(pi), wherein A (*) indicates the area of band of position *, such as rij<
t2, then it is assumed that Large Construction vehicle V at this timeiArea saturation degree it is inadequate, judge simultaneously otherwise to raise arm to be non-labeled as raising arm at this time.
The described FasterRCNN model steps for collecting pictures training detection Large Construction vehicle are：
21) picture of Large Construction vehicle is collected；
22) picture frame mark is carried out to the Large Construction vehicle in picture；
23) parameters such as iterations, learning rate are set and carry out the training of Faster RCNN algorithms, and by trained mould
Type is stored.
The described FasterRCNN model steps for collecting pictures training detection Large Construction automobile body are：
31) picture of Large Construction vehicle is collected；
23) picture frame mark is carried out to the Large Construction automobile body in picture；
33) parameters such as iterations, learning rate are set and carry out the training of Faster RCNN algorithms, and by trained mould
Type is stored.
It is described use Faster RCNN algorithms detection Large Construction vehicle step for：
41) the FasterRCNN moulds of the Large Constructions vehicles such as trained detection crane, excavator, cement pump truck are loaded
Type；
42) FasterRCNN model inspection Large Construction vehicles are used to input picture, obtains Large Construction vehicle Vi's
Classification li and band of position pi (xistart, yistart, xiend, yiend).
It is described use Faster RCNN algorithms detection Large Construction automobile body step for：
51) the FasterRCNN models of trained detection Large Construction automobile body are loaded；
52) FasterRCNN model inspection Large Construction automobile bodies are used to input picture, obtains Large Construction vehicle
The classification blj and band of position bpj (xjstart, yjstart, xjend, yjend) of vehicle body VBj.
The present invention carries out Large Construction vehicle identification in single image and further identifies that it raises arm behavior, with other methods
It compares, mainly has the advantage that：
1. carrying out the identification of Large Construction vehicle and main body car body, recognition accuracy higher using Faster RCNN；
2. further being identified according to vehicle entirety and intersubjective overlapping relation and raising arm behavior, practicability is stronger.
This method identifies raising for Large Construction vehicle for realizing in line protection region based on image procossing
Arm behavior, Detection accuracy are high.
Description of the drawings
Fig. 1 is that the medium-and-large-sized Construction traffic of the present invention raises arm detection algorithm flow chart.
Specific implementation mode
For an image, gives Large Construction vehicle and raise arm detection example.Method with reference to the present invention is detailed
Illustrate the specific steps that the example is implemented, it is as follows：
(1) training that collects pictures detects the FasterRCNN models of Large Construction vehicle；The Large Construction vehicle packet
Include crane, excavator, cement pump truck etc.；Step is：
21) picture of Large Construction vehicle is collected；
22) picture frame mark is carried out to the Large Construction vehicle in picture；
23) parameters such as iterations, learning rate are set and carry out the training of Faster RCNN algorithms, and by trained mould
Type is stored.
(2) training that collects pictures detects the FasterRCNN models of Large Construction automobile body；It is specific as follows：
31) picture of Large Construction vehicle is collected；
23) picture frame mark is carried out to the Large Construction automobile body in picture；
33) parameters such as iterations, learning rate are set and carry out the training of Faster RCNN algorithms, and by trained mould
Type is stored.
(3) Faster RCNN algorithms are used to detect Large Construction vehicle；It is specific as follows：
41) the FasterRCNN moulds of the Large Constructions vehicles such as trained detection crane, excavator, cement pump truck are loaded
Type；
42) FasterRCNN model inspection Large Construction vehicles are used to input picture, obtains Large Construction vehicle Vi's
Classification li and band of position pi (xistart, yistart, xiend, yiend).
(4) Faster RCNN algorithms are used to detect Large Construction automobile body；It is specific as follows：
51) the FasterRCNN models of trained detection Large Construction automobile body are loaded；
52) FasterRCNN model inspection Large Construction automobile bodies are used to input picture, obtains Large Construction vehicle
The classification blj and band of position bpj (xjstart, yjstart, xjend, yjend) of vehicle body VBj.
(5) ratio for calculating Large Construction automobile body area and entire vehicle area, is then judged to raising less than certain threshold value
Arm is then labeled, and is not otherwise marked；The specific steps are：
1) a Large Construction vehicle V identified is chosen respectivelyiThe Large Construction automobile body VB identified with onej
2) judge ViAnd VBjClassification liAnd bljIf they are unequal, step 1) is gone to, is otherwise carried out in next step
3) Large Construction vehicle V is calculatediBand of position piWith Large Construction automobile body VBjBand of position bpjPhase
Hand over region spij(xstart,ystart,xend,yend), computational methods spij(xstart,ystart,xend,yend)=(max (xi start,
xj start),max(yi start,yj start),min(xi end,xj end),min(yi end,yj end)), such as xend<xstartOr yend<ystart, then
Indicate ViBand of position piWith VBjBand of position bpjThere is no intersection, goes to step 1) at this time；
4) vehicle body area including degree c is calculatedij=A (spij)/A(bpj), wherein A (*) indicates the area of band of position *, such as
cij<t1, then it is assumed that Large Construction vehicle V at this timeiWith Large Construction automobile body VBjIncluding degree it is inadequate, be not belonging to same
Vehicle turns to step 1) at this time；
Calculate vehicle body area coincidence factor rij=A (spij)/A(pi), wherein A (*) indicates the area of band of position *, such as rij<
t2, then it is assumed that Large Construction vehicle V at this timeiArea saturation degree it is inadequate, judge simultaneously otherwise to raise arm to be non-labeled as raising arm at this time.
Because of vehicle body area coincidence factor<0.5, therefore it is determined as that crane raises arm.
The present invention identifies raising for Large Construction vehicle for realizing in line protection region based on image procossing
Arm behavior, Detection accuracy are high.
Claims (6)
1. a kind of Large Construction vehicle raises arm detection algorithm, it is characterised in that include the following steps：
(1) training that collects pictures detects the FasterRCNN models of Large Construction vehicle；
(2) training that collects pictures detects the FasterRCNN models of Large Construction automobile body；
(3) Faster RCNN algorithms are used to detect Large Construction vehicle；
(4) Faster RCNN algorithms are used to detect Large Construction automobile body；
(5) ratio for calculating Large Construction automobile body area and entire vehicle area, is then judged to raising arm then less than certain threshold value
It is labeled, does not otherwise mark；The specific steps are：
1) a Large Construction vehicle V identified is chosen respectivelyiThe Large Construction automobile body VB identified with onej
2) judge ViAnd VBjClassification liAnd bljIf they are unequal, step 1) is gone to, is otherwise carried out in next step
3) Large Construction vehicle V is calculatediBand of position piWith Large Construction automobile body VBjBand of position bpjIntersection
Domain spij(xstart,ystart,xend,yend), computational methods spij(xstart,ystart,xend,yend)=(max (xi start,
xj start),max(yi start,yj start),min(xi end,xj end),min(yi end,yj end)), such as xend<xstartOr yend<ystart, then
Indicate ViBand of position piWith VBjBand of position bpjThere is no intersection, goes to step 1) at this time；
4) vehicle body area including degree c is calculatedij=A (spij)/A(bpj), wherein A (*) indicates the area of band of position *, such as cij<
t1, then it is assumed that Large Construction vehicle V at this timeiWith Large Construction automobile body VBjIncluding degree it is inadequate, be not belonging to same vehicle, this
When turn to step 1)；
Calculate vehicle body area coincidence factor rij=A (spij)/A(pi), wherein A (*) indicates the area of band of position *, such as rij<t2, then
Think Large Construction vehicle V at this timeiArea saturation degree it is inadequate, judge simultaneously otherwise to raise arm to be non-labeled as raising arm at this time.
2. Large Construction vehicle according to claim 1 raises arm detection algorithm, it is characterised in that：The instruction that collects pictures
Practice detection Large Construction vehicle FasterRCNN model steps be：
21) picture of Large Construction vehicle is collected；
22) picture frame mark is carried out to the Large Construction vehicle in picture；
23) set the parameters such as iterations, learning rate carry out Faster RCNN algorithms training, and by trained model into
Row storage.
3. Large Construction vehicle according to claim 1 raises arm detection algorithm, it is characterised in that：The instruction that collects pictures
Practice detection Large Construction automobile body FasterRCNN model steps be：
31) picture of Large Construction vehicle is collected；
23) picture frame mark is carried out to the Large Construction automobile body in picture；
33) set the parameters such as iterations, learning rate carry out Faster RCNN algorithms training, and by trained model into
Row storage.
4. Large Construction vehicle according to claim 1 raises arm detection algorithm, it is characterised in that：Described uses Faster
RCNN algorithms detect Large Construction vehicle step：
41) the FasterRCNN models of the Large Constructions vehicles such as trained detection crane, excavator, cement pump truck are loaded；
42) FasterRCNN model inspection Large Construction vehicles are used to input picture, obtains the classification of Large Construction vehicle Vi
Li and band of position pi (xistart, yistart, xiend, yiend).
5. Large Construction vehicle according to claim 1 raises arm detection algorithm, it is characterised in that：Described uses Faster
RCNN algorithms detect Large Construction automobile body step：
51) the FasterRCNN models of trained detection Large Construction automobile body are loaded；
52) FasterRCNN model inspection Large Construction automobile bodies are used to input picture, obtains Large Construction automobile body
The classification blj and band of position bpj (xjstart, yjstart, xjend, yjend) of VBj.
6. Large Construction vehicle according to claim 1 raises arm detection algorithm, it is characterised in that：Large Construction vehicle includes
Crane, excavator, cement pump truck.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810608932.2A CN108805207B (en) | 2018-06-13 | 2018-06-13 | Method for detecting large construction vehicle lifting arm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810608932.2A CN108805207B (en) | 2018-06-13 | 2018-06-13 | Method for detecting large construction vehicle lifting arm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108805207A true CN108805207A (en) | 2018-11-13 |
| CN108805207B CN108805207B (en) | 2020-08-04 |
Family
ID=64085793
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810608932.2A Active CN108805207B (en) | 2018-06-13 | 2018-06-13 | Method for detecting large construction vehicle lifting arm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108805207B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102880881A (en) * | 2012-09-25 | 2013-01-16 | 常州大学 | Method for identifying car type on basis of binary support vector machines and genetic algorithm |
| CN105469062A (en) * | 2015-12-04 | 2016-04-06 | 电子科技大学 | Principal component analysis network framework based vehicle type identification method |
| CN105744232A (en) * | 2016-03-25 | 2016-07-06 | 南京第五十五所技术开发有限公司 | Method for preventing power transmission line from being externally broken through video based on behaviour analysis technology |
| CN105894701A (en) * | 2016-04-05 | 2016-08-24 | 江苏电力信息技术有限公司 | Large construction vehicle identification and alarm method for preventing external damage to transmission lines |
| CN106570477A (en) * | 2016-10-28 | 2017-04-19 | 中国科学院自动化研究所 | Vehicle model recognition model construction method based on depth learning and vehicle model recognition method based on depth learning |
| US9760806B1 (en) * | 2016-05-11 | 2017-09-12 | TCL Research America Inc. | Method and system for vision-centric deep-learning-based road situation analysis |
| CN108109385A (en) * | 2018-01-18 | 2018-06-01 | 南京杰迈视讯科技有限公司 | A kind of vehicle identification of power transmission line external force damage prevention and hazardous act judgement system and method |
-
2018
- 2018-06-13 CN CN201810608932.2A patent/CN108805207B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102880881A (en) * | 2012-09-25 | 2013-01-16 | 常州大学 | Method for identifying car type on basis of binary support vector machines and genetic algorithm |
| CN105469062A (en) * | 2015-12-04 | 2016-04-06 | 电子科技大学 | Principal component analysis network framework based vehicle type identification method |
| CN105744232A (en) * | 2016-03-25 | 2016-07-06 | 南京第五十五所技术开发有限公司 | Method for preventing power transmission line from being externally broken through video based on behaviour analysis technology |
| CN105894701A (en) * | 2016-04-05 | 2016-08-24 | 江苏电力信息技术有限公司 | Large construction vehicle identification and alarm method for preventing external damage to transmission lines |
| US9760806B1 (en) * | 2016-05-11 | 2017-09-12 | TCL Research America Inc. | Method and system for vision-centric deep-learning-based road situation analysis |
| CN106570477A (en) * | 2016-10-28 | 2017-04-19 | 中国科学院自动化研究所 | Vehicle model recognition model construction method based on depth learning and vehicle model recognition method based on depth learning |
| CN108109385A (en) * | 2018-01-18 | 2018-06-01 | 南京杰迈视讯科技有限公司 | A kind of vehicle identification of power transmission line external force damage prevention and hazardous act judgement system and method |
Non-Patent Citations (2)
| Title |
|---|
| 丁健配: ""基于图像差异的输电线路周边隐患检测系统"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 王林 等: ""Faster R-CNN 模型在车辆检测中的应用"", 《计算机应用》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108805207B (en) | 2020-08-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11840239B2 (en) | Multiple exposure event determination | |
| CN108537197B (en) | Lane line detection early warning device and method based on deep learning | |
| CN109460699B (en) | Driver safety belt wearing identification method based on deep learning | |
| Philipsen et al. | Traffic light detection: A learning algorithm and evaluations on challenging dataset | |
| Farag et al. | Road lane-lines detection in real-time for advanced driving assistance systems | |
| WO2020103893A1 (en) | Lane line property detection method, device, electronic apparatus, and readable storage medium | |
| CN111209777A (en) | Lane line detection method and device, electronic device and readable storage medium | |
| CN110852313B (en) | Parking space detection method | |
| Liu et al. | Vision-based real-time lane marking detection and tracking | |
| CN107545263A (en) | A kind of object detecting method and device | |
| WO2020181426A1 (en) | Lane line detection method and device, mobile platform, and storage medium | |
| CN111259707B (en) | Training method of special linear lane line detection model | |
| CN109508659A (en) | A kind of face identification system and method for crossing | |
| CN111967396A (en) | Processing method, device and equipment for obstacle detection and storage medium | |
| CN107705577A (en) | A kind of real-time detection method and system based on lane line demarcation vehicle peccancy lane change | |
| CN111967384A (en) | Vehicle information processing method, device, equipment and computer readable storage medium | |
| CN111178181B (en) | Traffic scene segmentation method and related device | |
| CN108446581A (en) | A kind of unmanned plane detection method in adverse circumstances | |
| Ozgunalp | Combination of the symmetrical local threshold and the sobel edge detector for lane feature extraction | |
| CN117351298A (en) | Mine operation vehicle detection method and system based on deep learning | |
| CN108805207A (en) | A kind of Large Construction vehicle raises arm detection algorithm | |
| CN110634120B (en) | Vehicle damage judgment method and device | |
| CN111144361A (en) | Road lane detection method based on binaryzation CGAN network | |
| Ben Romdhane et al. | A lane detection and tracking method for driver assistance system | |
| TWI619099B (en) | Intelligent multifunctional driving assisted driving recording method and 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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 210024 civil defense building, No. 189, Guangzhou Road, Gulou District, Nanjing, Jiangsu Applicant after: Jiangsu Electric Power Information Technology Co., Ltd. Applicant after: STATE GRID JIANGSU ELECTRIC POWER Co. Address before: 210024 civil defense building, No. 189, Guangzhou Road, Gulou District, Nanjing, Jiangsu Applicant before: Jiangsu Electric Power Information Technology Co., Ltd. Applicant before: STATE GRID JIANGSU ELECTRIC POWER COMPANY |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |