CN117480979A - Deep learning-based intelligent under-film tobacco seedling hole breaking and residual film recycling device and method - Google Patents
Deep learning-based intelligent under-film tobacco seedling hole breaking and residual film recycling device and method Download PDFInfo
- Publication number
- CN117480979A CN117480979A CN202311479030.0A CN202311479030A CN117480979A CN 117480979 A CN117480979 A CN 117480979A CN 202311479030 A CN202311479030 A CN 202311479030A CN 117480979 A CN117480979 A CN 117480979A
- Authority
- CN
- China
- Prior art keywords
- tobacco
- breaking
- residual film
- film
- deep learning
- 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
- 241000208125 Nicotiana Species 0.000 title claims abstract description 106
- 235000002637 Nicotiana tabacum Nutrition 0.000 title claims abstract description 106
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000004064 recycling Methods 0.000 title claims abstract description 20
- 238000013135 deep learning Methods 0.000 title claims abstract description 16
- 238000003860 storage Methods 0.000 claims abstract description 23
- 230000003993 interaction Effects 0.000 claims description 10
- 238000001514 detection method Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 7
- 238000011084 recovery Methods 0.000 claims description 7
- 238000004891 communication Methods 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 5
- 238000012544 monitoring process Methods 0.000 claims description 5
- 230000002708 enhancing effect Effects 0.000 claims description 2
- 230000004807 localization Effects 0.000 claims description 2
- 239000012528 membrane Substances 0.000 abstract description 16
- 230000007613 environmental effect Effects 0.000 abstract description 5
- 230000010365 information processing Effects 0.000 abstract 1
- 230000008569 process Effects 0.000 description 23
- 230000000694 effects Effects 0.000 description 11
- 239000002985 plastic film Substances 0.000 description 5
- 229920006255 plastic film Polymers 0.000 description 5
- 238000004140 cleaning Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000005286 illumination Methods 0.000 description 4
- 238000005457 optimization Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 241000196324 Embryophyta Species 0.000 description 3
- 239000003814 drug Substances 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 239000012535 impurity Substances 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 description 2
- 208000002474 Tinea Diseases 0.000 description 2
- 241000130764 Tinea Species 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000005520 cutting process Methods 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000011049 filling Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000001965 increasing effect Effects 0.000 description 2
- 229910052744 lithium Inorganic materials 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000002362 mulch Substances 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- 238000012827 research and development Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 208000012260 Accidental injury Diseases 0.000 description 1
- 206010056377 Bone tuberculosis Diseases 0.000 description 1
- 206010007247 Carbuncle Diseases 0.000 description 1
- 206010017553 Furuncle Diseases 0.000 description 1
- 241000238631 Hexapoda Species 0.000 description 1
- 206010062717 Increased upper airway secretion Diseases 0.000 description 1
- 208000009360 Osteoarticular Tuberculosis Diseases 0.000 description 1
- 208000004078 Snake Bites Diseases 0.000 description 1
- 206010043866 Tinea capitis Diseases 0.000 description 1
- 241000607479 Yersinia pestis Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000012271 agricultural production Methods 0.000 description 1
- 230000009418 agronomic effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 206010003246 arthritis Diseases 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001684 chronic effect Effects 0.000 description 1
- 235000019504 cigarettes Nutrition 0.000 description 1
- 230000003749 cleanliness Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001784 detoxification Methods 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000008014 freezing Effects 0.000 description 1
- 238000007710 freezing Methods 0.000 description 1
- 241000411851 herbal medicine Species 0.000 description 1
- 208000015181 infectious disease Diseases 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 238000013140 knowledge distillation Methods 0.000 description 1
- 208000037805 labour Diseases 0.000 description 1
- 238000011068 loading method Methods 0.000 description 1
- 238000011070 membrane recovery Methods 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- VYQNWZOUAUKGHI-UHFFFAOYSA-N monobenzone Chemical compound C1=CC(O)=CC=C1OCC1=CC=CC=C1 VYQNWZOUAUKGHI-UHFFFAOYSA-N 0.000 description 1
- 235000015097 nutrients Nutrition 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 208000026435 phlegm Diseases 0.000 description 1
- 239000011148 porous material Substances 0.000 description 1
- 230000036544 posture Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000001681 protective effect Effects 0.000 description 1
- 238000013138 pruning Methods 0.000 description 1
- 238000010223 real-time analysis Methods 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 239000004575 stone Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 231100000611 venom Toxicity 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01G—HORTICULTURE; CULTIVATION OF VEGETABLES, FLOWERS, RICE, FRUIT, VINES, HOPS OR SEAWEED; FORESTRY; WATERING
- A01G13/00—Protecting plants
- A01G13/02—Protective coverings for plants; Coverings for the ground; Devices for laying-out or removing coverings
- A01G13/0256—Ground coverings
- A01G13/0287—Devices for laying-out or removing ground coverings
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/50—Constructional details
Landscapes
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Toxicology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Environmental Sciences (AREA)
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Catching Or Destruction (AREA)
Abstract
The invention relates to a deep learning-based device and a method for intelligently breaking holes and recycling residual films of tobacco seedlings under films. Can shoot the high definition picture of tobacco seedling under the membrane in real time through the industry camera at the steady in-process that evenly goes forward of dolly to corresponding conveying is for lightweight marginal equipment, and marginal equipment can confirm the accurate position of tobacco seedling under the membrane after carrying out information processing, simultaneously gives the broken hole device of accurate position information feedback to laser, and then control laser carries out suitable broken hole operation to the membrane that covers on the tobacco seedling. After the hole breaking operation is completed, the residual film can be adsorbed into the storage box by the negative pressure rotary screen storage box at the rear of the trolley in time, so that the high-efficiency recycling of the residual film of the tobacco seedlings is realized, and the field environmental protection of the tobacco seedling growth is ensured.
Description
Technical Field
The invention belongs to the field of intelligent agriculture, and particularly relates to an intelligent perforation and residual film recycling device and method for tobacco seedlings under films.
Background
Tobacco is a commercial crop that is widely cultivated worldwide and has a very long history. The leaves can be used for making cigarettes and other products. Tobacco planting and processing has become a worldwide important industry, and it has important implications for rural economic development and farmer income increase. The tobacco yield in China is rich, the market demand is increasing, and large-scale tobacco planting drives employment and economic growth in rural areas, so that the tobacco planting method becomes an important way and stable income source for peasants. In addition, the development of the tobacco industry also drives the development of the industrial chains of sales, logistics, processing, research and development and the like in the related fields, and forms a complete industrial system. Tobacco has economic value and medicinal value not only can be ignored. According to the records of Chinese herbal medicine assembly, the tobacco Wen Weigan has the effects of detumescence, detoxification, disinfestation and the like, is mainly used for treating diseases such as furuncle, pyogenic infections, tinea capitis, tinea alba, tinea falciparum, venomous snake bite and the like, and can also treat diseases such as carbuncle on the back, wind phlegm, bone tuberculosis, chronic suppurative knee arthritis and the like. Can also be used for killing four pests, insects and the like. The chemical components in tobacco play an important role in the fields of medicine research and development, plant medicine, traditional Chinese medicine formula and the like.
In recent years, the technology of underfilm planting has been widely used in tobacco production. However, during the growth of tobacco under the film, it is necessary to ensure a sufficient supply of oxygen and nutrients. Therefore, three to four days later after the tobacco seedlings are covered with the film, hole breaking operation is needed to adjust the temperature and humidity under the film so as to promote the normal growth of the tobacco seedlings. At present, the most traditional hole breaking method adopts manual work, farmers judge the positions of tobacco seedlings under the film by naked eyes, and perform corresponding mechanical hole breaking operation; in this way, farmers have long working time and are easy to fatigue, and the broken holes of the instruments possibly damage the leaves, stems and the like of tobacco seedlings, and a series of problems of Kong Kongda small inconsistency, inaccurate hole positions, low efficiency and the like exist; meanwhile, if the broken residual film cannot be recovered in time, the corresponding field environmental protection problem can be generated. Therefore, there is a strong need for a smart method and apparatus for under-film tobacco perforation and residual film recovery. At present, artificial intelligence rapidly develops, and the mode of manually carrying out tobacco film subsurface hole breaking and residual film recovery can not meet corresponding agricultural production requirements, so that the research on the device for film subsurface tobacco intelligent hole breaking and residual film recovery based on deep learning and laser has very important practical significance.
Disclosure of Invention
The invention aims to provide a smart hole-breaking and residual film recycling device for tobacco under a film, which aims to solve the key problems of large manual hole-breaking workload, difficult residual film recycling and low efficiency.
In order to achieve the above purpose, the technical scheme adopted is as follows: broken hole of tobacco wisdom and incomplete membrane recovery unit include:
the device include laser device, industry camera, lightweight marginal equipment and rotatory plastic film residue containing box of negative pressure, wherein, industry camera is connected through internet access's mode and lightweight marginal equipment, and laser device and lightweight marginal equipment are connected through the mode of electric wire, and plastic film residue containing box is fixed through mechanical connection's mode and device structure.
Further, the laser device also comprises a safety sensor and a safety protection system which are arranged on the laser device, and the safety sensor and the laser device are connected in a circuit mode so as to monitor the operation condition of the trolley in real time and ensure the operation safety.
Further, the device also comprises a man-machine interaction system, and network connection is carried out between the man-machine interaction device and the lightweight edge device, so that the safety and convenience of operation are ensured.
Further, rotatory plastic film residue containing box of negative pressure in be equipped with intelligent sensor, intelligent sensor passes through the mode of circuit and links to each other with lightweight marginal equipment to realize retrieving the intelligent monitoring of plastic film residue.
In yet another aspect, a deep learning-based method for intelligently breaking holes and recovering residual films of tobacco seedlings under a film, which is realized by the device according to the claims 1-4, comprises the following steps:
the industrial camera captures image data and,
the identification and localization is performed by means of a target detection algorithm,
the intelligent hole breaking is realized through a laser device,
and recycling the residual film by rotating the residual film storage box under negative pressure.
Further, the method also comprises the step of enhancing and optimizing the data set while collecting the data so as to improve the generalization capability and the robustness of the model.
Further, the method further comprises feedback correction of the laser hole breaking position so as to improve the hole breaking accuracy.
Further, the method further comprises uploading the hole breaking result and the key data to a big data cloud platform for data sharing and communication.
Further, the method also comprises the step of monitoring in real time through a safety protection system to protect the working environment and personnel.
Further, the method also comprises the steps of setting operation parameters, processing faults and checking data through a man-machine interaction system, and performing remote communication control with the lightweight edge equipment
The invention has the beneficial effects that
The device suitable for membrane rupture of the tobacco seedlings transplanted under the membrane can save more than 60 labors and greatly improve the speed and the precision of membrane rupture; compared with the traditional mechanical seedling picking process, the laser film breaking process has the advantages that mechanical damage such as breaking can be avoided, and the mechanical seedling picking process has the characteristics of accuracy and high efficiency.
According to the technical scheme, the method and the device can realize that the position of tobacco under the film is identified and determined by using an industrial camera in the process of stably and uniformly advancing the trolley, a circle of hole burning is performed through two horizontally symmetrical lasers in an ingenious mode, and finally, the residual film left after the hole breaking is sucked into the storage box through the negative pressure rotary screen for recycling, so that the accuracy and the high efficiency of operation and the cleanliness of the field environment are ensured. The invention brings remarkable technical progress and economic benefit to the field of tobacco planting and processing, and plays a positive promoting role in environmental protection and sustainable development in the tobacco production process.
Drawings
FIG. 1 is a perspective view of the whole structure of the present invention
FIG. 2 is a view of the parts of the whole part of the present invention
FIG. 3 is a diagram of an industrial camera structure according to the present invention
FIG. 4 is a diagram showing the construction of a laser device according to the present invention
FIG. 5 is a diagram showing the structure of a negative pressure rotary storage box according to the present invention
FIG. 6 is a block diagram of a lightweight edge device according to the present invention
FIG. 7 is an overall flow chart of the present invention
FIG. 8 is a main line drawing of the present invention
FIG. 9 is a flow chart of the under-film tobacco seedling hole breaking process according to the invention
FIG. 10 is a flow chart of a big data cloud platform of the present invention
FIG. 11 is a flow chart of the ridge line determination of the present invention
FIG. 12 is a flow chart of the security protection of the present invention
1, a left laser; 2. a right laser; 3. an industrial camera; 4. negative pressure rotary storage box; 5. lightweight edge devices.
Detailed Description
The invention is further described with reference to the drawings and the flow chart.
Because the plant spacing of the tobacco seedlings is smaller, and the plastic film is covered above the tobacco seedlings, and the tobacco seedlings are identified and positioned in the uniform speed of the intelligent trolley, the tobacco seedlings are required to be identified and positioned from the background (soil, sand and stone and weeds) quickly and accurately, and the tobacco seedlings cannot be damaged in the process of identifying the tobacco seedlings and carrying out hole breaking operation. In the practical operation process, the lightweight edge equipment has the characteristics of convenience, rapidness and flexibility, and is more suitable for being used as core control equipment compared with a traditional computer or industrial personal computer.
As shown in fig. 2, each part of the carriage is a laser device 1 and 2, an industrial camera 3, a negative pressure rotary film residue storage box 4, and a lightweight edge device 5.
The deep learning visual recognition system is used for rapidly and accurately detecting, recognizing and positioning the young tobacco seedlings below the mulching film, and is a precondition and key for automatically cutting and breaking the mulching film above the tobacco seedlings to break the film of the tobacco seedlings. And the detailed information of the tobacco position under the mulch is obtained through a deep learning YOLOV8 algorithm, so that the position of the young tobacco seedling under the mulch can be rapidly and accurately identified and positioned. The high-precision camera is used for collecting images, so that clear and high-quality film tobacco seedling images can be obtained under various illumination conditions. In the running process of the trolley, a large number of data sets of the tobacco seedlings under the film are acquired in real time through the camera, and the image data sets of the tobacco seedlings under the film with different postures, illumination conditions and shielding conditions are included, so that the robustness and the accuracy of the recognition model are improved. The collected film tobacco seedling image is preprocessed, including denoising, image enhancement, edge detection and other operations, so that the accuracy and efficiency of subsequent image processing are improved. Through iterative training and optimization, the accuracy and generalization capability of the model are improved, the specific accurate position of tobacco seedlings under the membrane is identified and positioned in real time, and the position information is transmitted to a laser membrane breaking control system.
The laser hole breaking device has the characteristics of high power and high stability, and has smaller spot diameter and high focusing capacity so as to ensure the hole breaking precision and effect. The laser forms high temperature points on the film around the tobacco by focusing the laser beam, thereby achieving hole breaking. And the two lasers are symmetrically arranged on the horizontal rod, and a gear transmission system is arranged at the position where the horizontal rod is connected with the lasers and is used for controlling the horizontal movement of the lasers. By synchronizing with the movement of the trolley, the two symmetrical lasers can move horizontally in the process of steady advancing at a constant speed at the same time, and a complete semicircular arc is drawn. The specific process is as follows: before reaching the specific position of the tobacco seedling under the film, the two symmetrical lasers start from the center point of the horizontal rod at the same time, respectively move horizontally leftwards and horizontally rightwards on the rod, the initial speed is far greater than the advancing speed of the trolley, and then the speed is uniformly reduced until the speed is reduced to zero when the tobacco seedling is at the left side and the right side; then the uniform acceleration movement is started until the speed is increased to the maximum just in front of the center position of the tobacco seedling, and when the two lasers return to the initial position and are attached together, the speed is instantaneously braked to become zero. According to the circumferential speed and mechanics principle, two lasers skillfully draw two semicircular arcs in the process, and because the two lasers finally return to the center point of the horizontal rod at the same time and the starting point and the ending point are at the same position, the two lasers are combined together to form a complete circular arc shape, so that the shape of a hole can be completely drawn at the accurate position of a tobacco seedling film. By reasonable laser parameter settings, such as power, pulse frequency, focusing point position of laser beam and the like, accurate control in the hole breaking operation process is better realized. Meanwhile, the lower end of the laser device is provided with a spherical structure, so that free rotation in space can be realized, the position and the angle during laser emission can be controlled better, and the membrane rupture effect can be guaranteed more.
Spherical structure with adjustable angle: the junction design of laser instrument is spherical structure, can realize the free rotation in space, and this angle that makes the laser instrument can adjust when laser emission to adapt to the broken hole demand of different tobacco seedling ponds' position and shape. Through the rotation angle of control globular structure, carry out accurate location to draw the circular arc in the exact position, ensure the accuracy of membrane hole position and shape, realize finer and nimble broken hole operation.
And a residual film recovery and cleaning system. After the hole is broken, a small part of residual film is left, and if the residual film cannot be degraded in the future, the environment is polluted, so that the residual film after the hole is broken is recovered in time, and the method has very good practical significance. Therefore, the negative pressure rotary screen film suction device is designed and positioned at the rear of the bottom of the trolley and has negative pressure suction and rotary screen functions. After the hole breaking operation, the residual film is sucked into the storage box by negative pressure suction, the volume of the residual film is compressed through rotary screening, and meanwhile, impurities are discharged, so that the environmental cleaning and protection of the tobacco field are realized. Meanwhile, the intelligent sensor is further arranged in the storage box, so that the filling condition of the residual film in the storage box can be monitored, when the residual film in the storage box is too much, corresponding signals can be sent out, and the residual film and impurities in the storage box are cleaned regularly, so that the normal operation and the efficient operation of the system are maintained.
Data recording and analyzing system: the lightweight edge equipment is provided with a related data recording device, and can record key data parameters of each hole breaking operation and residual film recovery, so that the hole breaking effect and the residual film recovery performance are further known, the subsequent data analysis and optimization are facilitated, and the system is adjusted and improved in real time.
The system for updating the iteration and model of the broken Kong Suanfa is built in light-weight edge equipment, based on analysis of tobacco seedling morphology and position information, an industrial camera monitors the position and the film breaking effect of tobacco seedlings under a film in real time, and then the light-weight edge equipment performs real-time analysis and error correction. Through feedback adjustment of Kalman filtering, according to the error of the previous n times, the accurate position and key parameters in the (n+1) th time of hole breaking are automatically corrected, an upgrade iteration is carried out on a hole breaking algorithm, and the model important parameters of the YOLO target detection network are updated in real time, so that the accuracy and stability of hole breaking operation are improved, and the method is suitable for changing environments and different film breaking scenes.
Energy supply system: in order to ensure continuous operation of the system, a reliable energy supply system is required, using a stable and reliable lithium battery pack as an energy source. The lithium battery is a battery type with high energy density, long service life and portability, and is widely applied to the fields of mobile equipment, medical equipment, electric tools and the like. And the trolley is low-temperature resistant, is not afraid of freezing in winter, has stable performance and strong power, and can better ensure continuous and stable power supply of the trolley during field operation.
Human-computer interaction interface and safety protection system: multiple safety protection measures are arranged in the system to ensure the safety of operators and surrounding environment. The laser membrane rupture system adopts a plurality of safety protection measures. The safety sensor is arranged at the proper position of the trolley, so that the approach of a human body can be detected in time and the laser can be stopped in time, and the accidental injury can be avoided. In addition, the system is also provided with an emergency stop device and corresponding protective measures, and by monitoring the real-time running condition of the trolley, once an abnormal condition occurs, the system can immediately stop operation and send out a warning signal, so that the safety of operators is ensured. In addition, the system provides a friendly man-machine interaction interface, and operators can conveniently set parameters, process faults and view data through the lightweight edge equipment, so that the safety and convenience of operation are ensured.
Big data sharing and cooperation cloud platform: and the big data cloud platform is formed into a data sharing and cooperation platform, so that information sharing and cooperation in the agricultural field are promoted. Farmers, experts, research institutions and the like can upload own tobacco seedling membrane breaking data to the platform and share and communicate with experience of other users so as to promote technical innovation and improve membrane breaking technology level and farmland management effect of tobacco seedlings of different varieties together. Due to the flexibility and expandability of the big data cloud platform, the big data cloud platform has strong computing capacity and a large amount of stored resources, and can smoothly carry out the optimization work of the self-adaptive algorithm. The error sources between the arc positions of the laser film breaking and the actual positions of the tobacco seedlings are found out by continuously collecting and analyzing large-scale tobacco seedling data sets from a plurality of operation scenes, algorithm iteration and model updating are continuously carried out, so that the tobacco seedling growth environments under different planting environments and illumination conditions are adapted, and the film breaking accuracy and stability are improved.
First some preparations in the early stage:
step one: the industrial camera shoots images in the process of uniform speed advancing of the trolley, and acquires image data sets of tobacco seedlings and tobacco seedlings in the pond from different angles and different dimensions to serve as comparison. The method comprises the steps of obtaining a large number of images in various environments such as sunny days, cloudy days, rainy days and the like, wherein the images comprise a series of information such as space positions, colors, shape textures and the like, and simulating the real visual field condition of an industrial camera in the operation process of the trolley.
Step two: preprocessing is carried out after a large number of data sets are acquired, corresponding data enhancement is carried out on the collected pictures, a series of operations including random cutting, size transformation, embedding, noise addition, random brightness and contrast adjustment and the like are carried out, the data sets are further expanded, the risk of overfitting is reduced, and the generalization capability and the robustness of the model are enhanced. And (3) manually marking the accurate positions of the tobacco seedlings under the film in the image, and constructing a training set, a verification set and a test set according to the proportion of 8:1:1, wherein each data set contains the same number and proportion of various samples.
Step three: training the collected sub-membranous tobacco seedling data set by adopting a target detection YOLOV8 algorithm, and simultaneously carrying out knowledge distillation, model pruning and migration learning to lighten the model, compress the parameter quantity and reduce the calculated quantity. And iterating for many times to optimize network parameters, further obtaining better detection performance, comparing target detection and model identification effects after training the model file, continuously modifying the model structure, and optimizing the program. And comprehensively evaluating through evaluation indexes MAP, FPS, parameters, model size, F-score, detection speed and the like, selecting a sub-film tobacco seedling recognition Model with optimal effect, robustness and accuracy, transplanting and loading the sub-film tobacco seedling recognition Model into lightweight edge equipment, and preparing for precise recognition in real operation.
Then specific workflow details:
the overall flow is as shown in flow chart 7: after the system is initialized, the intelligent trolley forwards and stably forwards and uniformly forwards along the tobacco seedling ridge line as shown in fig. 1, and the industrial camera acquires the tobacco seedling image under the film in real time as shown in fig. 3 and transmits the image data to the light edge equipment;
as shown in fig. 6, the lightweight edge device judges whether there is a tobacco seedling under the film by an industrial camera, if there is no tobacco seedling, a candidate frame cannot be generated, and at this time, missing seedling information is recorded and the next recognition is performed.
If tobacco seedlings exist, a tobacco seedling target detection frame is generated at the corresponding position, so that the position of the tobacco seedlings under the film is accurately positioned, the light-weight edge equipment can transmit the position information of the target detection frame to the laser device through a relevant protocol, and the laser device is controlled to draw a circle of circular arc on the periphery of the tobacco seedling film, so that intelligent hole breaking operation is performed.
Meanwhile, the edge equipment can carry out feedback correction on the position of the broken hole through Kalman filtering control, so that the result is more and more accurate.
After the operation of the current ridge line is completed, the trolley turns to enter a new ridge line for operation.
Meanwhile, an intelligent safety protection system is arranged in the lightweight edge equipment, and safety conditions in operation are monitored in real time.
If the potential safety hazard exists, stopping all operations to ensure the safety of operators.
And finally, after the operation is finished, uploading the film breaking effect and key data of the whole land block to a big data cloud platform by the lightweight edge equipment, and sharing and exchanging data to jointly promote the iterative upgrading of the tobacco seedling hole breaking agronomic technology.
S1, an intelligent hole breaking process of a camera and laser is shown in FIG. 9: in the process of the trolley advancing forward at a constant speed and steadily, the industrial camera shoots images of tobacco seedlings under the film in real time and feeds the images back to the lightweight edge equipment; and judging the situation of tobacco seedlings under the film by the edge equipment.
And recording missing seedling information without tobacco seedlings, and identifying tobacco seedlings in the next pond.
If tobacco seedlings exist, the positions of the tobacco seedlings under the film are accurately positioned, and the edge equipment controls the laser to draw a circle of circular arc along the periphery of the tobacco seedlings, so that hole breaking operation is skillfully performed.
The laser hole breaking detailed process has the characteristics of high power and high stability, and the two lasers which are symmetrically arranged on the horizontal bar have smaller spot diameter and high focusing capacity so as to ensure the hole breaking precision and effect.
The laser, as shown in fig. 4, forms a high temperature spot on the film around the tobacco by focusing the laser beam, thereby achieving a hole breaking. And the two lasers are symmetrically arranged on the horizontal rod, and a gear transmission system is arranged at the position where the horizontal rod is connected with the lasers and is used for controlling the horizontal movement of the lasers. By synchronizing with the movement of the trolley, the two symmetrical lasers can move horizontally in the process of steady advancing at a constant speed at the same time, and a complete semicircular arc is drawn.
The specific process is as follows: before reaching the specific position of the tobacco seedling under the film, the two symmetrical lasers start from the center point of the horizontal rod at the same time, respectively move horizontally leftwards and horizontally rightwards on the rod, the initial speed is far greater than the advancing speed of the trolley, and then the speed is uniformly reduced until the speed is reduced to zero when the tobacco seedling is at the left side and the right side; then the uniform acceleration movement is started, the speed is maximized when the position right in front of the center of the tobacco seedling is reached, and the speed is instantaneously braked to become zero when the two lasers return to the initial position and are attached together. According to the circumferential speed and mechanics principle, two lasers skillfully draw two semicircular arcs in the process, and because the two lasers finally return to the center point of the horizontal rod at the same time and the starting point and the ending point are at the same position, the two lasers are combined together to form a complete circular arc shape, so that the shape of a hole can be completely drawn at the accurate position on a tobacco seedling film.
By reasonable laser parameter settings, such as power, pulse frequency, focusing point position of laser beam and the like, accurate control in the hole breaking operation process is better realized. Meanwhile, the lower end of the laser device is provided with a spherical structure, so that free rotation in space can be realized, the position and the angle during laser emission can be controlled better, and the membrane rupture effect can be guaranteed more.
S2, the feedback correction system, the edge equipment can analyze the errors of the previous n times of arc positions and the tobacco seedling positions, and adjust and correct key parameters in the errors through Kalman filtering control, so that the (n+1) th time of arc positioning errors are gradually reduced, the hole breaking accuracy is higher and higher, and full preparation is made for the next accurate positioning.
S3, a big data cloud platform is shown in a flow chart 10: the cloud platform is a platform for sharing and cooperating data. Farmers, experts, research institutions and the like can upload the film breaking data of the vegetation seedlings to the cloud platform for experience sharing and communication so as to promote technical innovation and jointly improve the film breaking technology and farmland management level.
The big data cloud platform has strong computing power and a large amount of storage resources, and can flexibly develop the optimization work of the self-adaptive algorithm. By collecting and analyzing large-scale tobacco seedling data sets from a plurality of operation scenes, finding out error sources between the arc positions of laser film breaking and the actual positions of tobacco seedlings, continuously carrying out algorithm iteration and model updating so as to adapt to tobacco seedling growth environments under different planting environments and illumination conditions and improve the accuracy and stability of film breaking
S4, updating the operation ridge line, as shown in a flow chart 11: when the trolley works, if the camera shoots that a pond exists under the front membrane, the operation is continued after the ridge line is not finished;
when no pond is arranged in front of the camera, the edge equipment can control the trolley to turn at a proper position at the tail end of the ridge line, so that the trolley enters a new ridge line for operation.
S5, a safety protection system is shown in FIG. 12: multiple safety protection measures are arranged in the system to ensure the safety of operators and surrounding environment.
The laser is provided with a safety sensor, and when the hole breaking operation is carried out, the lightweight edge equipment monitors the operation condition of the trolley in real time.
If potential safety hazards occur, for example, when a technician is too close to the laser during operation, the laser power is too high or the potential safety hazards exist in the movement gesture of the trolley, the safety protection system is triggered, the operation of the trolley is stopped in an emergency mode, and a warning signal is sent out to ensure the safety of the operator.
S6, a man-machine interaction system: a friendly man-machine interaction interface is designed on the lightweight edge equipment, and farmers or related personnel can conveniently perform parameter setting, fault processing and data viewing through the platform interface, so that the safety and convenience of operation are ensured.
Meanwhile, the device can be remotely communicated with lightweight edge equipment through a computer or an industrial personal computer, so that the process of membrane rupture operation can be conveniently monitored, controlled and adjusted remotely, and the instantaneity and accuracy of the pore rupture operation are ensured. S7, a residual film recycling and cleaning system: after the hole is broken, a small part of residual film is left, and if the residual film cannot be degraded in future days, the environment is polluted. Therefore, the residual film after hole breaking is timely recovered, and the method has very important practical significance.
The negative pressure rotary screen film suction device is positioned at the rear bottom of the trolley as shown in fig. 5, and has negative pressure suction and rotary screen functions. After the hole breaking operation, the residual film is sucked into the storage box by negative pressure suction, and is separated and collected in the box by rotary screening, and meanwhile, impurities are discharged, so that the environmental cleaning and protection of the tobacco field are realized. Meanwhile, the intelligent sensor is arranged in the storage box, so that the filling condition of the residual film in the storage box can be monitored, and a timely feedback signal is provided, so that the residual film in the storage box is cleaned regularly, and the normal operation and the efficient operation of the system are maintained.
Those skilled in the art will appreciate that implementing all or part of the above-described methods in accordance with the embodiments may be accomplished by way of a computer program stored on a computer readable storage medium, which when executed may comprise the steps of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ReadOnlyMemory, ROM) or a random access memory (RandomABBessMemory, RAM).
It should be understood that the detailed description of the technical solution of the present invention, given by way of preferred embodiments, is illustrative and not restrictive. Modifications of the technical solutions described in the embodiments or equivalent substitutions of some technical features thereof may be performed by those skilled in the art on the basis of the present description; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. The device comprises a laser device, an industrial camera, light-weight edge equipment and a negative-pressure rotary residual film storage box, wherein the industrial camera is connected with the light-weight edge equipment through a network connection mode, the laser device and the light-weight edge equipment are connected through a wire mode, and the residual film storage box is fixed with a device structure through a mechanical connection mode.
2. The deep learning-based intelligent under-film tobacco seedling hole breaking and residual film recycling device according to claim 1, wherein the laser device also comprises a safety sensor and a safety protection system which are arranged on the laser device, and the safety sensor and the laser device are connected in a line manner so as to monitor the operation condition of the trolley in real time and ensure the operation safety.
3. The deep learning-based intelligent hole breaking and residual film recycling device for tobacco seedlings under films, which is characterized by further comprising a man-machine interaction system, wherein the man-machine interaction system is connected with lightweight edge equipment through a network, so that the safety and convenience of operation are guaranteed.
4. The deep learning-based intelligent under-film tobacco seedling hole breaking and residual film recycling device is characterized in that an intelligent sensor is arranged in a negative pressure rotary residual film storage box and is connected with lightweight edge equipment in a line mode, so that intelligent monitoring of recycling residual films is achieved.
5. A deep learning-based method for intelligently breaking holes and recycling residual films of tobacco seedlings under films, which is realized by the device according to the claims 1-4, and is characterized by comprising the following steps:
the industrial camera captures image data and,
the identification and localization is performed by means of a target detection algorithm,
the intelligent hole breaking is realized through a laser device,
and recycling the residual film by rotating the residual film storage box under negative pressure.
6. The deep learning-based method for intelligently breaking holes and recycling residual films of tobacco seedlings under films according to claim 5, which is characterized by further comprising the step of enhancing and optimizing a data set while collecting data so as to improve generalization capability and robustness of a model.
7. The method for intelligently breaking holes and recycling residual films on basis of deep learning of tobacco seedlings under films according to claim 5 or 6, wherein the method further comprises feedback correction of the positions of laser hole breaking so as to improve the hole breaking accuracy.
8. The deep learning-based intelligent hole breaking and residual film recycling method for tobacco seedlings under films, which is characterized by further comprising uploading hole breaking results and key data to a big data cloud platform for data sharing and communication.
9. The deep learning-based method for intelligently breaking holes and recycling residual films of tobacco seedlings under films according to claims 5-8, which is characterized by further comprising the step of monitoring in real time through a safety protection system to protect the working environment and personnel.
10. The deep learning-based intelligent perforation and residual film recovery method for tobacco seedlings under films according to claims 5-9, further comprising the steps of setting operation parameters, processing faults and checking data through a man-machine interaction system, and performing remote communication control with lightweight edge equipment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311479030.0A CN117480979A (en) | 2023-11-08 | 2023-11-08 | Deep learning-based intelligent under-film tobacco seedling hole breaking and residual film recycling device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311479030.0A CN117480979A (en) | 2023-11-08 | 2023-11-08 | Deep learning-based intelligent under-film tobacco seedling hole breaking and residual film recycling device and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117480979A true CN117480979A (en) | 2024-02-02 |
Family
ID=89670398
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311479030.0A Pending CN117480979A (en) | 2023-11-08 | 2023-11-08 | Deep learning-based intelligent under-film tobacco seedling hole breaking and residual film recycling device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117480979A (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2017024028A (en) * | 2015-07-17 | 2017-02-02 | トヨタ自動車株式会社 | Laser drilling device |
| KR200485559Y1 (en) * | 2017-06-05 | 2018-01-25 | 김동오 | Punching Apparatus For Mulching Film For Agriculture |
| CN107750780A (en) * | 2017-12-04 | 2018-03-06 | 山东科技大学 | A kind of crawler type peanut mulch film rupture of membranes draws seedling machine and its method of work |
| CN109479590A (en) * | 2018-12-27 | 2019-03-19 | 宁波市基业金属材料有限公司 | A kind of broken residual film secondary recovery processing mechanism of small-sized incomplete film-recovering machine |
| CN209120612U (en) * | 2018-11-09 | 2019-07-19 | 通辽市农业科学研究院 | Sand ground drought rice hole-ground film fabric machine |
| CN112894961A (en) * | 2021-02-22 | 2021-06-04 | 陈明星 | Agricultural planting is with broken membrane machine of drawing seedling based on artifical intelligent control of AI |
| CN114467593A (en) * | 2022-03-16 | 2022-05-13 | 安徽省农业科学院烟草研究所 | Film-coating and hole-punching integrated equipment for flue-cured tobacco |
| WO2022258653A1 (en) * | 2021-06-10 | 2022-12-15 | Eto Magnetic Gmbh | Device for detecting a sprouting of sown seeds, agricultural sensor device, and agricultural monitoring and/or control method and system |
| CN115885749A (en) * | 2022-12-13 | 2023-04-04 | 徐州艾博生物科技有限公司 | Plastic film punching intelligent robot for plant planting |
| CN116806609A (en) * | 2023-02-06 | 2023-09-29 | 合肥工业大学 | Membrane rupture method for peanut planting and membrane rupture assembly matched with membrane rupture method |
-
2023
- 2023-11-08 CN CN202311479030.0A patent/CN117480979A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2017024028A (en) * | 2015-07-17 | 2017-02-02 | トヨタ自動車株式会社 | Laser drilling device |
| KR200485559Y1 (en) * | 2017-06-05 | 2018-01-25 | 김동오 | Punching Apparatus For Mulching Film For Agriculture |
| CN107750780A (en) * | 2017-12-04 | 2018-03-06 | 山东科技大学 | A kind of crawler type peanut mulch film rupture of membranes draws seedling machine and its method of work |
| CN209120612U (en) * | 2018-11-09 | 2019-07-19 | 通辽市农业科学研究院 | Sand ground drought rice hole-ground film fabric machine |
| CN109479590A (en) * | 2018-12-27 | 2019-03-19 | 宁波市基业金属材料有限公司 | A kind of broken residual film secondary recovery processing mechanism of small-sized incomplete film-recovering machine |
| CN112894961A (en) * | 2021-02-22 | 2021-06-04 | 陈明星 | Agricultural planting is with broken membrane machine of drawing seedling based on artifical intelligent control of AI |
| WO2022258653A1 (en) * | 2021-06-10 | 2022-12-15 | Eto Magnetic Gmbh | Device for detecting a sprouting of sown seeds, agricultural sensor device, and agricultural monitoring and/or control method and system |
| CN114467593A (en) * | 2022-03-16 | 2022-05-13 | 安徽省农业科学院烟草研究所 | Film-coating and hole-punching integrated equipment for flue-cured tobacco |
| CN115885749A (en) * | 2022-12-13 | 2023-04-04 | 徐州艾博生物科技有限公司 | Plastic film punching intelligent robot for plant planting |
| CN116806609A (en) * | 2023-02-06 | 2023-09-29 | 合肥工业大学 | Membrane rupture method for peanut planting and membrane rupture assembly matched with membrane rupture method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xie et al. | Actuators and sensors for application in agricultural robots: A review | |
| CN109191074A (en) | Wisdom orchard planting management system | |
| CN109240304A (en) | A kind of precision planting system and method | |
| JP2018108040A (en) | Work machine, control device, and program for control | |
| CN113597874B (en) | Weeding robot and weeding path planning method, device and medium thereof | |
| JP6616378B2 (en) | Information processing system and program | |
| Selvi et al. | Weed detection in agricultural fields using deep learning process | |
| WO2023204243A1 (en) | Forestry management system and forestry management method | |
| CN107121535A (en) | A kind of effective crop growth environment monitoring system | |
| Ji et al. | Research on key technology of truss tomato harvesting robot in greenhouse | |
| Wang et al. | A review on the application of computer vision and machine learning in the tea industry | |
| Wang et al. | Research advance in phenotype detection robots for agriculture and forestry | |
| WO2022091092A1 (en) | System and method for indoor crop management | |
| CN211628101U (en) | Football court lawn maintenance system | |
| CN117480979A (en) | Deep learning-based intelligent under-film tobacco seedling hole breaking and residual film recycling device and method | |
| CN115756037A (en) | Greenhouse vegetable management and control method and system | |
| Lei et al. | Technologies and Equipment of Mechanized Blossom Thinning in Orchards: A Review | |
| Sharma | Recognition of Anthracnose Injuries on Apple Surfaces using YOLOV 3-Dense | |
| Kushwaha | Robotic and mechatronic application in agriculture | |
| CN114882365A (en) | Intelligent monitoring system for peony grafting | |
| JP6616377B2 (en) | Information processing system and program | |
| CN112699805A (en) | Intelligent recognition system for vegetable pest control | |
| CN214175117U (en) | Crop disease and pest recognition device based on it is embedded | |
| JP2015223100A (en) | Information processing device, plant body management system, state management work detection method and plant body management method | |
| CN108633533B (en) | Tree automatic tending method and device, electronic equipment and storage medium |
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 |