CN112308922A - Deep learning-based automatic goods-searching positioning method and system for shuttle - Google Patents
Deep learning-based automatic goods-searching positioning method and system for shuttle Download PDFInfo
- Publication number
- CN112308922A CN112308922A CN202011336937.8A CN202011336937A CN112308922A CN 112308922 A CN112308922 A CN 112308922A CN 202011336937 A CN202011336937 A CN 202011336937A CN 112308922 A CN112308922 A CN 112308922A
- Authority
- CN
- China
- Prior art keywords
- network model
- shuttle
- target detection
- images
- searching
- 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
- 238000013135 deep learning Methods 0.000 title claims abstract description 16
- 238000000034 method Methods 0.000 title claims abstract description 15
- 238000001514 detection method Methods 0.000 claims abstract description 25
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 15
- 238000012549 training Methods 0.000 claims abstract description 12
- 238000004519 manufacturing process Methods 0.000 claims abstract description 4
- 238000012545 processing Methods 0.000 claims description 10
- SAZUGELZHZOXHB-UHFFFAOYSA-N acecarbromal Chemical compound CCC(Br)(CC)C(=O)NC(=O)NC(C)=O SAZUGELZHZOXHB-UHFFFAOYSA-N 0.000 claims description 6
- 238000011176 pooling Methods 0.000 claims description 5
- 238000010586 diagram Methods 0.000 claims description 3
- 230000003137 locomotive effect Effects 0.000 claims description 2
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013138 pruning Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30248—Vehicle exterior or interior
- G06T2207/30252—Vehicle exterior; Vicinity of vehicle
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a deep learning-based automatic goods-searching and positioning method and system for a shuttle car, wherein the method comprises the following steps of collecting images of various trays, marking coordinates of the trays in the images, and manufacturing a tray training sample set; inputting the tray training sample set into a built combined convolutional neural network model, wherein the combined convolutional neural network model comprises a classification network and a target detection network, and training after setting hyper-parameters; step three, storing the weights of the trained classification network model and the target detection network model; and step four, classifying and identifying the images acquired by the shuttle vehicle by using the classification network model and the target detection network model, controlling the rotating speed of a driving motor of the shuttle vehicle, and accurately parking and searching goods. The invention realizes automatic goods-searching and positioning of the shuttle vehicle by means of an image recognition target detection technology, improves the robustness of a shuttle vehicle positioning system, reduces the workload of field debugging personnel and ensures the accurate positioning of the shuttle vehicle.
Description
Technical Field
The invention relates to the technical field of intelligent logistics and deep learning computer vision, in particular to an automatic goods-searching positioning method and system for a shuttle vehicle based on deep learning.
Background
With the rapid development of national economy, the automatic stereoscopic warehouse is widely applied in various industries and becomes an important component of modern logistics systems. The shuttle car as an important component in the stereoscopic warehouse has the characteristics of convenience and high efficiency in goods conveying. However, the existing shuttle car has the defects of consuming time and labor in early installation and debugging, and the fundamental reason of the existing shuttle car is that the technology of adopting the infrared sensor to search goods and position cannot adapt to tray structures of different styles. Usually, the sensing distance and the angle of the infrared sensor are adjusted by field debugging personnel, and even the infrared sensor cannot be accurately positioned on the tray adopting the hollow pattern, so that the shuttle car needs a set of goods searching and positioning system with stronger robustness.
Disclosure of Invention
The invention aims to provide an automatic goods-searching positioning method and system for a shuttle car based on deep learning, and aims to solve the problems of low robustness and high labor consumption of positioning of the shuttle car in the existing logistics industry by adopting an infrared sensor.
In order to achieve the purpose, the technical scheme of the invention is as follows:
an automatic goods-searching and positioning method for a shuttle car based on deep learning comprises the following steps:
collecting images of various trays, marking tray coordinates in the images, and manufacturing a tray training sample set;
inputting the tray training sample set into a constructed and combined convolutional neural network model, wherein the convolutional neural network model comprises a classification network and a target detection network, and training after setting hyper-parameters;
step three, storing the weights of the trained classification network model and the target detection network model;
and step four, classifying and identifying the images acquired by the shuttle vehicle by using the classification network model and the target detection network model, controlling the rotating speed of a driving motor of the shuttle vehicle, and accurately parking and searching goods.
Further, the classification network model adopts a DenseNet convolutional neural network to automatically extract image features, and then uses a full-connection layer or a global pooling layer to output a classification result so as to judge whether the acquired image contains a tray.
Further, the target detection network model processes images including trays detected in the classification network, a convolutional neural network is adopted to automatically extract a characteristic diagram, and a fast RCNN target detection framework is adopted.
An automatic goods-searching and positioning system of a shuttle car based on deep learning comprises a high-definition camera, an image processing unit and a shuttle car driving motor, wherein the camera and the shuttle car driving motor are connected with the image processing unit, and the image processing unit comprises a graphic processor, and a classification network model and a target detection network model which are embedded in the graphic processor; the high-definition camera is installed on the shuttle car and used for collecting images and transmitting the images to the graphic processor.
Furthermore, the number of the high-definition cameras is two, and the high-definition cameras are obliquely and upwards installed at the head and the parking spaces of the shuttle car.
According to the shuttle vehicle automatic goods-searching positioning method and system based on deep learning, the shuttle vehicle automatic goods-searching positioning is realized by means of an image recognition target detection technology, the robustness of a shuttle vehicle positioning system is improved, the workload of field debugging personnel is reduced, and the accurate positioning of the shuttle vehicle is ensured.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description, serve to explain the principles of the invention, and are best understood by reference to the accompanying drawings in which:
fig. 1 is a flow chart of an automatic cargo finding and positioning method of a shuttle vehicle based on deep learning according to the invention.
Detailed Description
The technical solution of the present invention is further described in detail with reference to the accompanying drawings and examples.
As shown in fig. 1, the automatic goods-searching and positioning method for the shuttle car based on deep learning of the present invention includes the following steps:
collecting images of various trays, marking tray coordinates in the images, and manufacturing a tray training sample set;
inputting the tray training sample set into a built combined convolutional neural network model, wherein the combined convolutional neural network model comprises a classification network and a target detection network, and training after setting hyper-parameters;
step three, storing the weights of the trained classification network model and the trained target detection network model, and carrying out quantization and pruning to improve the model inference speed;
and step four, classifying and identifying the images acquired by the shuttle vehicle by using the classification network model and the target detection network model, controlling the rotating speed of a driving motor of the shuttle vehicle, and accurately parking and searching goods.
The classification network model adopts a DenseNet convolutional neural network to automatically extract image characteristics, and then uses a full-connection layer or a global pooling layer to output a classification result so as to judge whether the acquired image contains a tray. And processing images including the tray in the classification network by using a target detection network model, automatically extracting a characteristic diagram by using a convolutional neural network, and adopting a Faster RCNN target detection framework.
DenseNet is a convolutional neural network with dense connection, and the network greatly reduces the parameter quantity of the network and improves the repeated utilization rate of characteristic parameters by arranging bypass connection in the network.
The fast RCNN object detection framework comprises: the convolutional layer is used for extracting a feature map, the RPN (Region pro-social network) is used for generating a Region candidate box, the Roi Pooling (Region of Interests Pooling) is used for obtaining a feature map with uniform size, and the classification layer is used for classifying the recognition target.
An automatic goods-searching and positioning system of a shuttle car based on deep learning comprises a high-definition camera, an image processing unit and a shuttle car driving motor, wherein the camera and the shuttle car driving motor are connected with the image processing unit, and the image processing unit comprises a graphic processor, and a classification network model and a target detection network model which are embedded in the graphic processor; the image processor can be simultaneously connected with a plurality of high-definition cameras, the shuttle vehicle driving motor controls the rotating speed according to the operation result of the image processor, and the high-definition cameras are installed on the shuttle vehicle and used for collecting images and transmitting the images to the image processor.
In this embodiment, high definition digtal camera is two, and high definition digtal camera installs locomotive and the parking stall at the shuttle in the slant. Simultaneously, every high definition digtal camera is furnished with auxiliary lighting lamp, ensures to gather clear image under the abominable condition of ambient light.
The shuttle vehicle driving motor controls the rotating speed of the driving motor according to the operation result of the graphic processor, and accurate goods searching and positioning of the shuttle vehicle are ensured. When the classification network detects that the tray appears, the shuttle driving motor controls the shuttle to decelerate, and when the target detection network identifies the tray boundary, the shuttle is controlled to stop accurately.
The above-mentioned embodiments, objects, technical solutions and advantages of the present invention are further described in detail, it should be understood that the above-mentioned embodiments are only specific embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (5)
1. An automatic goods-searching and positioning method for a shuttle car based on deep learning is characterized by comprising the following steps:
collecting images of various trays, marking tray coordinates in the images, and manufacturing a tray training sample set;
inputting the tray training sample set into a constructed and combined convolutional neural network model, wherein the convolutional neural network model comprises a classification network and a target detection network, and training after setting hyper-parameters;
step three, storing the weights of the trained classification network model and the target detection network model;
and step four, classifying and identifying the images acquired by the shuttle vehicle by using the classification network model and the target detection network model, controlling the rotating speed of a driving motor of the shuttle vehicle, and accurately parking and searching goods.
2. The deep learning based shuttle car automatic cargo locating method according to claim 1, characterized in that: the classification network model adopts a DenseNet convolutional neural network to automatically extract image characteristics, and then uses a full-connection layer or a global pooling layer to output a classification result so as to judge whether the acquired image contains a tray.
3. The deep learning based shuttle car automatic cargo locating method according to claim 2, characterized in that: the target detection network model processes images including trays detected in the classification network, a convolutional neural network is adopted to automatically extract a characteristic diagram, and a Faster RCNN target detection framework is adopted.
4. The utility model provides an automatic goods positioning system that seeks of shuttle based on degree of depth study which characterized in that: the system comprises a high-definition camera, an image processing unit and a shuttle vehicle driving motor, wherein the camera and the shuttle vehicle driving motor are connected with the image processing unit, and the image processing unit comprises an image processor, a classification network model and a target detection network model which are embedded in the image processor; the high-definition camera is installed on the shuttle car and used for collecting images and transmitting the images to the graphic processor.
5. The deep learning based shuttle car auto-locating system of claim 4, wherein: the high definition digtal camera is two, and the high definition digtal camera is installed in the locomotive and the parking stall of shuttle car in the slant, every the high definition digtal camera is furnished with supplementary light for improve light luminance.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011336937.8A CN112308922A (en) | 2020-11-25 | 2020-11-25 | Deep learning-based automatic goods-searching positioning method and system for shuttle |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011336937.8A CN112308922A (en) | 2020-11-25 | 2020-11-25 | Deep learning-based automatic goods-searching positioning method and system for shuttle |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112308922A true CN112308922A (en) | 2021-02-02 |
Family
ID=74335519
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011336937.8A Pending CN112308922A (en) | 2020-11-25 | 2020-11-25 | Deep learning-based automatic goods-searching positioning method and system for shuttle |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112308922A (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106845424A (en) * | 2017-01-24 | 2017-06-13 | 南京大学 | Road surface remnant object detection method based on depth convolutional network |
CN107967468A (en) * | 2018-01-19 | 2018-04-27 | 刘至键 | A kind of supplementary controlled system based on pilotless automobile |
CN108875911A (en) * | 2018-05-25 | 2018-11-23 | 同济大学 | One kind is parked position detecting method |
CN109466725A (en) * | 2018-10-11 | 2019-03-15 | 重庆邮电大学 | It is a kind of that system is salvaged based on the intelligent floater of neural network and image recognition |
CN109870983A (en) * | 2017-12-04 | 2019-06-11 | 北京京东尚科信息技术有限公司 | Handle the method, apparatus of pallet stacking image and the system for picking of storing in a warehouse |
CN110109456A (en) * | 2019-04-25 | 2019-08-09 | 杭州师范大学钱江学院 | A kind of trolley automatic Pilot method |
CN111898430A (en) * | 2020-06-25 | 2020-11-06 | 北京电通慧梯物联网科技有限公司 | Elevator car personnel intelligent identification method based on deep learning |
-
2020
- 2020-11-25 CN CN202011336937.8A patent/CN112308922A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106845424A (en) * | 2017-01-24 | 2017-06-13 | 南京大学 | Road surface remnant object detection method based on depth convolutional network |
CN109870983A (en) * | 2017-12-04 | 2019-06-11 | 北京京东尚科信息技术有限公司 | Handle the method, apparatus of pallet stacking image and the system for picking of storing in a warehouse |
CN107967468A (en) * | 2018-01-19 | 2018-04-27 | 刘至键 | A kind of supplementary controlled system based on pilotless automobile |
CN108875911A (en) * | 2018-05-25 | 2018-11-23 | 同济大学 | One kind is parked position detecting method |
CN109466725A (en) * | 2018-10-11 | 2019-03-15 | 重庆邮电大学 | It is a kind of that system is salvaged based on the intelligent floater of neural network and image recognition |
CN110109456A (en) * | 2019-04-25 | 2019-08-09 | 杭州师范大学钱江学院 | A kind of trolley automatic Pilot method |
CN111898430A (en) * | 2020-06-25 | 2020-11-06 | 北京电通慧梯物联网科技有限公司 | Elevator car personnel intelligent identification method based on deep learning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11720799B2 (en) | Object detection neural networks | |
CN111007073B (en) | Method and system for online detection of part defects in additive manufacturing process | |
CN110222702B (en) | Industrial vehicle with dome lamp based positioning | |
US9174830B1 (en) | Industrial vehicles with point fix based localization | |
US11227434B2 (en) | Map constructing apparatus and map constructing method | |
CN110780305A (en) | Track cone bucket detection and target point tracking method based on multi-line laser radar | |
CN108873904B (en) | Unmanned parking method and device for mining vehicle and readable storage medium | |
CN110766979A (en) | Parking space detection method for automatic driving vehicle | |
CN111198496A (en) | Target following robot and following method | |
CN114488194A (en) | Method for detecting and identifying targets under structured road of intelligent driving vehicle | |
US20200133283A1 (en) | Proceedable direction detection apparatus and proceedable direction detection method | |
Zelener et al. | Cnn-based object segmentation in urban lidar with missing points | |
CN113850799A (en) | YOLOv 5-based trace DNA extraction workstation workpiece detection method | |
CN114863376A (en) | Road marking segmentation method and system based on vehicle-mounted laser point cloud data | |
US20240051146A1 (en) | Autonomous solar installation using artificial intelligence | |
CN112308922A (en) | Deep learning-based automatic goods-searching positioning method and system for shuttle | |
CN109635692B (en) | Scene re-identification method based on ultrasonic sensor | |
CN116935356A (en) | Weak supervision-based automatic driving multi-mode picture and point cloud instance segmentation method | |
CN103559492A (en) | Car logo recognition device and method | |
CN113917453A (en) | Multi-sensor fusion method based on radar and video | |
CN116071358B (en) | Automobile chassis detection method and readable storage medium | |
CN111860123B (en) | Method for identifying boundary of working area | |
CN117789521A (en) | Two-stage parking space line detection method for plane parking space | |
CN116883823A (en) | Underwater docking guide lamp identification and positioning method for underwater unmanned aircraft | |
Wong et al. | Parking lot management application using instance segmentation |
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 |