CN105230224A - Intelligent weeding machine and weed removing method thereof - Google Patents
Intelligent weeding machine and weed removing method thereof Download PDFInfo
- Publication number
- CN105230224A CN105230224A CN201510746468.XA CN201510746468A CN105230224A CN 105230224 A CN105230224 A CN 105230224A CN 201510746468 A CN201510746468 A CN 201510746468A CN 105230224 A CN105230224 A CN 105230224A
- Authority
- CN
- China
- Prior art keywords
- weeder
- weeds
- weeding
- integrated manipulator
- row
- 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
Landscapes
- Guiding Agricultural Machines (AREA)
- Image Processing (AREA)
Abstract
The invention provides an intelligent weeding machine and a weed removing method thereof. The intelligent weeding machine comprises a movable cart. The movable cart is provided with a working platform. A computer, cameras and an integrated controller are arranged on the working platform. The integrated controller is connected with the computer and the cameras in an electrical control mode, and an inter-row weeding device located on the front side and an in-row weeding device located on the back side are arranged below the working platform. The inter-row weeding device comprises a weeding mechanism, the in-row weeding device comprises two weeding mechanisms which are symmetrical in a left-and-right mode, an inside-outside moving mechanism is arranged between the two weeding mechanisms, the inside-outside moving mechanism comprises a transverse-movement air cylinder, and a stretchable rod of the transverse-movement air cylinder is connected with the weeding mechanism located on the outer side. The weeding mechanisms comprise upper comb plates and lower comb plates, cutters are arranged between the upper comb plates and the lower comb plates, in-row teeth are arranged at the front ends of the upper comb plates and the front ends of the lower comb plates, cutting edges are arranged at the edges of the front sides of the cutters, pesticide applying devices are arranged on the lower comb plates, and the integrated controller is connected with the transverse-movement air cylinder and the pesticide applying devices in an electrical control mode.
Description
Technical field
The invention belongs to agricultural machinery technological field, relate to a kind of farmland farming and weaving weeding apparatus, particularly a kind of intelligent weeder and weeds sweep-out method thereof.
Background technology
Weeds, with crops contention soil moisture and nutrient, occupy space, farmland, affect crop growth, not only reduce the seed output and quality of crops, also indirectly encouraged growing and spreading of damage by disease and insect.In order to management of weeds, develop various method weeding both at home and abroad, as artificial weeding, chemical weed control, weeding by machine, biological weed control, photochemistry weeding etc.Artificial weeding labour intensity is large, and inefficiency, has been not suitable with the needs of modern agricultural development.Chemical weed control is relatively time saving and energy saving, and weeding ratio is also higher, and for paroxysmal crop smothering, chemical weed control takes effect the fastest, so the dark welcome by vast farmers, becomes current main weeding system, is at home and abroad widely used.If but the improper use of chemical weed control weed killer herbicide, not only waste weed killer herbicide, and Practice for Pesticide Residue in Agricultural Products amount can be caused to exceed standard, the pesticide pollution such as soil, water environment, even personnel are poisoning.Therefore, while utilizing the effective controlling weeds of chemical herbicide, the amount of application of weed killer herbicide should be reduced to reduce its harm as far as possible.
At present, chemical weed control practices adopts plant equipment to realize the sprinkling of chemical herbicide usually.In China, usually adopt manual sprayer or engine-driven atomiser pesticide-applying weeding, also have aerial sprays weeding.In the U.S., current application equipment has boom sprayer, spread application utensil etc.But weeds and crop mix in farmland, no matter the shortcoming of these methods is that ground is with or without weeds, spraying herbicide equably all earthward, and spraying apparatus often has float of mist droplet phenomenon, so inevitably hurt crop during spraying herbicide, also too much can cause pesticide pollution because of weed killer herbicide consumption, thus the poisoning of weed killer herbicide occurs.Therefore, chemical weed control practices is being saved time and while labour, is also being brought the problem of waste and pollution, do not meet the requirement of modern agriculture.
In prior art, nearly all complete by artificial weeding or chemical weed control in the work of field weeding, labour intensity is large on the one hand, and efficiency is low; On the other hand, cause environmental deterioration, water quality, soil etc. pollute and are on the rise.Along with the shortage with between twenty and fifty labour that improves constantly of labor cost, and the needs of preserving the ecological environment, the intelligentized demand of weeding machineization is more and more stronger.
Summary of the invention
The object of the invention is to there are the problems referred to above for existing technology, propose and a kind ofly to cut off in the ranks separating with the weeds in row, adopt multiple RM to improve intelligent weeder and the weeds sweep-out method thereof of weed identification accuracy rate simultaneously.
Object of the present invention realizes by following technical proposal: a kind of intelligent weeder, comprise travelling car, described travelling car has workbench, described workbench is arranged calculator and integrated manipulator, the both sides of described calculator are respectively provided with camera, described integrated manipulator electricly connects calculator and camera, the below of described workbench arranges the weeder being in the ranks positioned at front side and the row weeder being positioned at rear side, described weeder in the ranks comprises a weeding mechanism, in described row, weeder comprises symmetrical Liang Bu weeding mechanism, between Liang Bu weeding mechanism, inside and outside travel mechanism is set, described inside and outside travel mechanism comprises flat-removing air cyclinder, the expansion link of described flat-removing air cyclinder connects the weeding mechanism being positioned at outside, described weeding mechanism comprises upper comb dent plate and lower fishback, between described upper comb dent plate and lower fishback, cutter is set, the front end of described upper comb dent plate and lower fishback arranges teeth in a row, the front side edge edge of described cutter is blade, described lower fishback arranges device for administration of drugs, described integrated manipulator electricly connects flat-removing air cyclinder and device for administration of drugs.
This intelligent weeder with drilling crops such as the veterinary antibiotics in agricultural leading industries for manipulating object.In row weeder left and right Liang Bu weeding mechanism enter in crop row carry out weeding operation time, it is positioned at the outer fix of the mechanism of weeder weeding in the ranks, and at once, the two ends distance of weeder is greater than the length of weeder in the ranks; And the left and right Liang Bu weeding mechanism of interior weeder of being expert at is when dodging crop, it is positioned at the inner side of the mechanism of weeder weeding in the ranks, and at once, the two ends distance of weeder is less than or equal to the length of weeder in the ranks.
In above-mentioned intelligent weeder, the blade indention shape of described cutter, the sawtooth wedge angle of described zigzag fashion is right angle or acute angle.
In above-mentioned intelligent weeder, the inner side of described lower fishback sets firmly ledger-plate, and described cutter connects movable mechanism.
In above-mentioned intelligent weeder, described movable mechanism comprises electromotor, the rotating shaft of described electromotor drives connecting eccentric wheel, described eccentric wheel is connected with connecting rod, the driving end of described connecting rod is hinged on above-mentioned cutter, corresponding formation with upper comb dent plate or lower fishback is slidably connected in the two sides up and down of described cutter, and described integrated manipulator electricly connects electromotor.Coordinated with the cutter of front and back reciprocating action by the ledger-plate of lower fishback, weeds are effectively cut.
In above-mentioned intelligent weeder, the teeth of described upper comb dent plate are downward-sloping, and its angle of inclination is 35 ° to 75 °, and the increment of described upper comb dent plate teeth is gathered close with the increment of lower fishback teeth mutually.The teeth of upper comb dent plate and lower fishback form Duckbill type thus, are beneficial to weeds combing further, to facilitate excision.
In above-mentioned intelligent weeder, described device for administration of drugs comprises deposits medicine-chest, described medicine-chest of depositing connects grug transportation tube, the import of described grug transportation tube arranges electrically-controlled valve, and the end of described grug transportation tube connects medicine dropper, and the teeth bottom side of described lower fishback arranges dispenser sponge, the mouth of pipe of described medicine dropper contacts above-mentioned dispenser sponge, place humidity sensor in described dispenser sponge, described humidity sensor connects above-mentioned integrated manipulator by circuit, and described integrated manipulator is electrically connected above-mentioned electrically-controlled valve.
In above-mentioned intelligent weeder, the bottom side of described workbench arranges fore-stock and after-poppet, the bottom of described fore-stock is fixedly mounted with the weeding mechanism of above-mentioned weeder in the ranks, described after-poppet has two sub-stents, the bottom of described sub-stent is fixedly mounted with the weeding mechanism of weeder in above-mentioned row, the bottom surface of described workbench sets firmly guide rail, and the top of described sub-stent arranges roller, and described roller embeds in guide rail to be formed and is slidably connected.Move flat in process inside and outside flat-removing air cyclinder driving weeding mechanism does, sub-stent carries out synchronization-sliding by roller in guide rail, by the guide effect of guide rail, promotes the stability of translation motion inside and outside weeding mechanism further.
In above-mentioned intelligent weeder, described travelling car also comprises four wheels, described four wheels are all connected to below workbench by double cross arm independent suspension, double cross arm independent suspension comprises Top Crossbeam and lower cross arm, Top Crossbeam is hinged on the bottom surface of workbench, vehicle wheel rotation is connected on lower cross arm, is hinged between Top Crossbeam and lower cross arm, between Top Crossbeam and lower cross arm, damper is housed.By the effect of damper, to avoid the walking vibrations caused because of pavement roughness, improve smooth running, to guarantee the stability of camera photographing picture.
A weeds sweep-out method for intelligent weeder, comprises the following steps:
1), by intelligent weeder be positioned over one end in the ranks, make the weeder in the ranks on front side of it be positioned in the ranks, in the row of rear side, weeder stretches in the adjacent inline crops in both sides;
2), intelligent weeder walks forward, is positioned at weeding mechanism in the ranks directly by upper comb dent plate and lower fishback combing weeds, then cuts off weeds by cutter, finally by device for administration of drugs to the weeds otch spread application just cut;
3), by the scene in the adjacent lines of camera shooting both sides, scene information is passed to integrated manipulator analysis, and analytic process and result is passed through Computer display out, comprise the method for following several identification crops:
I, color characteristic identification is utilized:
Point by point scanning is carried out to the image collected, splits green planting and soil according to the rgb value of each pixel, namely utilize color characteristic first from Soil Background, to isolate green planting, and then distinguish weeds and crop by the color characteristic detecting target area;
II, morphological feature identification is utilized:
Plants identification based on morphological feature realizes mainly through the border style (curvature or lobate feature) and whole blade shape (area, length, width, girth, square, non-dimensional ratio etc.) studying blade; The blade utilizing deformable template method to carry out identification division to block: the end first detecting blade, then model is placed on each end of blade, and by distortion, model being mated with whole blade, model structure is determined with the outline line representing blade shape by the variable skeleton of blade;
III, spectral signature identification is utilized:
Under the irradiation of visible ray (400 ~ 700nm wave band) or near infrared light (700 ~ 2500nm wave band), the spectral reflectivity of weeds and soil and weeds and crop also exists difference;
IV, textural characteristics identification is utilized:
Whether texture is the repeated arrangement of certain basic model (tone primitive), can occur significantly to change the border determined between two kinds of texture pattern by observing texture measure;
V, pattern-recongnition method:
Artificial neural network, SVMs, Bayes classifier, decision tree and discriminant analysis pattern is utilized to carry out identification weeds and crop;
4), identify and greenly plant for crop, integrated manipulator manipulation flat-removing air cyclinder retraction expansion link, drive weeding mechanism to move to the inside to realize dodging to leave inline crops, intelligence weeder continues to move ahead, until camera detection is completely through crop, integrated manipulator manipulates flat-removing air cyclinder again and stretches out expansion link, drives the movement laterally of weeding mechanism to reenter space in row;
5), identify green planting for weeds, stretch into row in weeding mechanism by upper comb dent plate and lower fishback combing weeds, then cut off weeds by cutter, finally by device for administration of drugs to the weeds otch spread application just cut.
In above-mentioned weeds sweep-out method, the humidity information of described humidity sensor sensing dispenser sponge, and humidity information is sent to integrated manipulator in real time, integrated manipulator carries out the judgement of dispenser sponge herb liquid content by detecting data, once judge that liquid is containing quantity not sufficient, integrated manipulator controls electrically-controlled valve and opens, and making liquid by depositing medicine-chest through grug transportation tube, dropping on dispenser sponge by medicine dropper; Until the sensing humidity value of humidity sensor is up to standard, integrated manipulator carries out dose judgement by detecting data, controls electrically-controlled valve further and closes, complete feeding operations.
Compared with prior art, this intelligent weeder and weeds sweep-out method thereof for research object, are divided into weeding two parts in weeding in the ranks and row with drilling crops such as the veterinary antibiotics in agricultural leading industries.Utilize the particularity of farmland weeding, the elevation information of weeds in crop plant and row is obtained based on binocular vision, by color characteristic identification ground blue target to reduce image Stereo matching operand, from horizontal pixel histogram-fitting curve, identify crop plant according to crop plant spatial distribution characteristic again, and then identify weeds in trip.In order to reduce the amount of calculation of Stereo matching, improving the real-time of image procossing, planning image background segmentation result and operating with Stereo matching and combine.Utilize the methods such as multispectral characteristic, shape facility, textural characteristics and pattern-recognition further, the special precision promoting weed identification in row, realizes the intelligent operation of automation field weeding.
Accompanying drawing explanation
Fig. 1 is the perspective view of this intelligent weeder.
Fig. 2 is the structural representation of weeder in row in this intelligent weeder.
Fig. 3 is the schematic diagram of cutter drives structure in this intelligent weeder.
In figure, 1, workbench; 2, camera; 3, calculator; 4, integrated manipulator; 5, fore-stock; 6, sub-stent; 7, upper comb dent plate; 8, lower fishback; 9, cutter; 10, eccentric wheel; 11, connecting rod; 12, flat-removing air cyclinder; 13, guide rail; 14, roller; 15, wheel.
Embodiment
Be below specific embodiments of the invention and by reference to the accompanying drawings, technical scheme of the present invention is further described, but the present invention be not limited to these embodiments.
As Fig. 1, shown in Fig. 2 and Fig. 3, this intelligent weeder comprises travelling car, travelling car has workbench 1, workbench 1 is arranged calculator 3 and integrated manipulator 4, the both sides of calculator 3 are respectively provided with camera 2, integrated manipulator 4 electricly connects calculator 3 and camera 2, the below of workbench 1 arranges the weeder being in the ranks positioned at front side and the row weeder being positioned at rear side, in the ranks weeder comprises a weeding mechanism, in row, weeder comprises symmetrical Liang Bu weeding mechanism, between Liang Bu weeding mechanism, inside and outside travel mechanism is set, inside and outside travel mechanism comprises flat-removing air cyclinder 12, the expansion link of flat-removing air cyclinder 12 connects the weeding mechanism being positioned at outside, weeding mechanism comprises upper comb dent plate 7 and lower fishback 8, between upper comb dent plate 7 and lower fishback 8, cutter 9 is set, the front end of upper comb dent plate 7 and lower fishback 8 arranges teeth in a row, the front side edge edge of cutter 9 is blade, lower fishback 8 arranges device for administration of drugs, integrated manipulator 4 electricly connects flat-removing air cyclinder 12 and device for administration of drugs.
In row weeder left and right Liang Bu weeding mechanism enter in crop row carry out weeding operation time, it is positioned at the outer fix of the mechanism of weeder weeding in the ranks, and at once, the two ends distance of weeder is greater than the length of weeder in the ranks; And the left and right Liang Bu weeding mechanism of interior weeder of being expert at is when dodging crop, it is positioned at the inner side of the mechanism of weeder weeding in the ranks, and at once, the two ends distance of weeder is less than or equal to the length of weeder in the ranks.
A kind of mode, the blade indention shape of cutter 9, the sawtooth wedge angle of zigzag fashion is right angle or acute angle.
Another kind of mode, the inner side of lower fishback 8 sets firmly ledger-plate, and cutter 9 connects movable mechanism.Movable mechanism comprises electromotor, the rotating shaft of electromotor drives connecting eccentric wheel 10, eccentric wheel 10 is connected with connecting rod 11, the driving end of connecting rod 11 is hinged on cutter 9, corresponding formation with upper comb dent plate 7 or lower fishback 8 is slidably connected in the two sides up and down of cutter 9, and integrated manipulator 4 electricly connects electromotor.Coordinated with the cutter 9 of front and back reciprocating action by the ledger-plate of lower fishback 8, weeds are effectively cut.
The teeth of upper comb dent plate 7 are downward-sloping, and its angle of inclination is 35 ° to 75 °, and the increment of upper comb dent plate 7 teeth is gathered close with the increment of lower fishback 8 teeth mutually.The teeth of upper comb dent plate 7 and lower fishback 8 form Duckbill type thus, are beneficial to weeds combing further, to facilitate excision.
Device for administration of drugs comprises deposits medicine-chest, deposit medicine-chest and connect grug transportation tube, the import of grug transportation tube arranges electrically-controlled valve, the end of grug transportation tube connects medicine dropper, the teeth bottom side of lower fishback 8 arranges dispenser sponge, and the mouth of pipe contact dispenser sponge of medicine dropper, places humidity sensor in dispenser sponge, humidity sensor connects integrated manipulator 4 by circuit, and integrated manipulator 4 is electrically connected electrically-controlled valve.
The bottom side of workbench 1 arranges fore-stock 5 and after-poppet, the bottom of fore-stock 5 is fixedly mounted with the weeding mechanism of weeder in the ranks, after-poppet has two sub-stents 6, the bottom of sub-stent 6 is fixedly mounted with the weeding mechanism of weeder in row, the bottom surface of workbench 1 sets firmly guide rail 13, the top of sub-stent 6 arranges roller 14, and roller 14 embeds in guide rail 13 to be formed and is slidably connected.Flat-removing air cyclinder 12 drive weeding mechanism do inside and outside move flat in process, sub-stent 6 carries out synchronization-sliding by roller 14 in guide rail 13, by the guide effect of guide rail 13, promotes further the stability of translation motion inside and outside weeding mechanism.
Travelling car also comprises four wheels 15, four wheels 15 are all connected to below workbench 1 by double cross arm independent suspension, double cross arm independent suspension comprises Top Crossbeam and lower cross arm, Top Crossbeam is hinged on the bottom surface of workbench 1, wheel 15 is rotatably connected on lower cross arm, be hinged between Top Crossbeam and lower cross arm, between Top Crossbeam and lower cross arm, damper be housed.By the effect of damper, to avoid the walking vibrations caused because of pavement roughness, improve smooth running, to guarantee the stability of camera 2 photographic picture.
A weeds sweep-out method for intelligent weeder, comprises the following steps:
1), by intelligent weeder be positioned over one end in the ranks, make the weeder in the ranks on front side of it be positioned in the ranks, in the row of rear side, weeder stretches in the adjacent inline crops in both sides;
2), intelligent weeder walks forward, is positioned at weeding mechanism in the ranks directly by upper comb dent plate 7 and lower fishback 8 combing weeds, then cuts off weeds by cutter 9, finally by device for administration of drugs to the weeds otch spread application just cut;
3), by camera 2 take the scene in the adjacent lines of both sides, scene information be passed to integrated manipulator 4 and analyze, and analytic process and result are shown by calculator 3, comprise the method for following several identification crops:
I, color characteristic identification is utilized:
Point by point scanning is carried out to the image collected, splits green planting and soil according to the rgb value of each pixel, namely utilize color characteristic first from Soil Background, to isolate green planting, and then distinguish weeds and crop by the color characteristic detecting target area;
Such as, research finds that the color of milpa blade is bottle green, and the color in strain heart region is light green color, to this, first utilize color characteristic from Soil Background, be partitioned into corn and weeds green plants region, recycling maize seedling strain heart color characteristic mapping maize plant, finally according to the connectivity of region identification milpa, the plant regional of the non-milpa obtained is corn weeds in field.
II, morphological feature identification is utilized:
Plants identification based on morphological feature realizes mainly through the border style (curvature or lobate feature) and whole blade shape (area, length, width, girth, square, non-dimensional ratio etc.) studying blade; The blade utilizing deformable template method to carry out identification division to block: the end first detecting blade, then model is placed on each end of blade, and by distortion, model being mated with whole blade, model structure is determined with the outline line representing blade shape by the variable skeleton of blade;
Such as, unifacial leaf and broadleaf weed difference in external form is very large, and monocotyledon weed blade is narrow and upright; Broadleaf weed blade is wide, and surface area is large.Therefore, the difference of form can be utilized to distinguish weeds and crop.Study the weed identification method based on morphological feature, come subregion tomato seedling and weeds by the elongation and tight ness rating analyzing blade.Utilize deformable template method to identify the weeds in Sugarbeet Fields, this model can remove the 83% weeds pixel be blocked, and weed identification rate is 81% ~ 87%.
III, spectral signature identification is utilized:
Under the irradiation of visible ray (400 ~ 700nm wave band) or near infrared light (700 ~ 2500nm wave band), the spectral reflectivity of weeds and soil and weeds and crop also exists difference;
Such as, use imaging spectrometer to determine beet and the reflectivity of weeds within the scope of 435 ~ 1000nm, result of study accuracy rate reaches 91%.Utilize quadratic discriminatory analysis method to choose the characteristic wavelength point identifying carrot and weeds, reach 72% with the discrimination of combination to carrot and weeds that 450,550 and 700nm are three interference filters of characteristic wavelength point.Gather the image of multi-wavelength at visible ray and near infrared band, then utilize two spectrum channels (660 and 800nm) from image background, split cotton and weeds, and develop the Robust Statistics algorithm based on local histogram.Measure the spectral reflectivities of Four Plants in 350 ~ 2500nm wavelength band such as cotton, field thistle, paddy rice and barnyard grass, discrimination is up to 100%.
IV, textural characteristics identification is utilized:
Whether texture is the repeated arrangement of certain basic model (tone primitive), can occur significantly to change the border determined between two kinds of texture pattern by observing texture measure;
Such as, utilize the textural characteristics of fourier spectrum to wheat and weeds to analyze, distinguish wheat and weeds with the difference of one direction and multi-direction texture pattern.Analytical calculation 10 gray level co-occurrence matrixes features, 12 gray enhance characteristic sum, 3 gray features, utilize these 25 textural characteristics to identify wheat weeds.With crop row center for benchmark chooses texture block, calculate and quantize the co-occurrence matrix that progression is the H color space of 8 grades, extract 5 textural characteristics parameters, utilize the classification of each piece of K means Method discriminant analysis.
V, pattern-recongnition method:
Artificial neural network, SVMs, Bayes classifier, decision tree and discriminant analysis pattern is utilized to carry out identification weeds and crop;
Such as, using the sunflower collected, weeds and soil map picture as inputting data modeling, construct an artificial neural network based on error backpropagation algorithm, and by reduce image resolution, cutting Soil Background edge to reduce the learning rate that picture size improves neutral net.The textural characteristics extracting corn and weeds, as the input vector of support vector machine classifier, utilizes this support vector machine method to identify corn and weeds, finds that the accuracy rate of support vector machine method identification weeds is higher than BP neutral net by contrast.With color and position feature for identification parameter, adopt Bayes classifier to distinguish cauliflower, weeds and soil, under the accurate a priori probability condition of acquisition, misclassification rate is minimum is 6.4%.Use discriminant analysis, decision tree and the recognition methods of artificial neural network Three models to identify the weeds in corn field respectively, weed identification rate is respectively 87%, 76% and 81%.
4), identify and greenly plant for crop, integrated manipulator 4 manipulates flat-removing air cyclinder 12 and to retract expansion link, drive weeding mechanism to move to the inside to realize dodging to leave inline crops, intelligence weeder continues to move ahead, until camera 2 detects completely through crop, integrated manipulator 4 manipulates flat-removing air cyclinder 12 again and stretches out expansion link, drives the movement laterally of weeding mechanism to reenter space in row;
5), identify green planting for weeds, stretch into row in weeding mechanism by upper comb dent plate 7 and lower fishback 8 combing weeds, then cut off weeds by cutter 9, finally by device for administration of drugs to the weeds otch spread application just cut.
The humidity information of humidity sensor sensing dispenser sponge, and humidity information is sent to integrated manipulator 4 in real time, integrated manipulator 4 carries out the judgement of dispenser sponge herb liquid content by detecting data, once judge that liquid is containing quantity not sufficient, integrated manipulator 4 controls electrically-controlled valve and opens, making liquid by depositing medicine-chest through grug transportation tube, dropping on dispenser sponge by medicine dropper; Until the sensing humidity value of humidity sensor is up to standard, integrated manipulator 4 carries out dose judgement by detecting data, controls electrically-controlled valve further and closes, complete feeding operations.
This intelligent weeder and weeds sweep-out method thereof for research object, are divided into weeding two parts in weeding in the ranks and row with drilling crops such as the veterinary antibiotics in agricultural leading industries.Utilize the particularity of farmland weeding, the elevation information of weeds in crop plant and row is obtained based on binocular vision, by color characteristic identification ground blue target to reduce image Stereo matching operand, from horizontal pixel histogram-fitting curve, identify crop plant according to crop plant spatial distribution characteristic again, and then identify weeds in trip.In order to reduce the amount of calculation of Stereo matching, improving the real-time of image procossing, planning image background segmentation result and operating with Stereo matching and combine.Utilize the methods such as multispectral characteristic, shape facility, textural characteristics and pattern-recognition further, the special precision promoting weed identification in row, realizes the intelligent operation of automation field weeding.
Specific embodiment described herein is only to the explanation for example of the present invention's spirit.Those skilled in the art can make various amendment or supplement or adopt similar mode to substitute to described specific embodiment, but can't depart from spirit of the present invention or surmount the scope that appended claims defines.
Although more employ workbench 1 herein; Camera 2; Calculator 3; Integrated manipulator 4; Fore-stock 5; Sub-stent 6; Upper comb dent plate 7; Lower fishback 8; Cutter 9; Eccentric wheel 10; Connecting rod 11; Flat-removing air cyclinder 12; Guide rail 13; Roller 14; Term such as wheel 15 grade, but do not get rid of the possibility using other term.These terms are used to be only used to describe and explain essence of the present invention more easily; The restriction that they are construed to any one additional is all contrary with spirit of the present invention.
Claims (10)
1. an intelligent weeder, comprise travelling car, described travelling car has workbench, described workbench is arranged calculator and integrated manipulator, the both sides of described calculator are respectively provided with camera, described integrated manipulator electricly connects calculator and camera, it is characterized in that, the below of described workbench arranges the weeder being in the ranks positioned at front side and the row weeder being positioned at rear side, described weeder in the ranks comprises a weeding mechanism, in described row, weeder comprises symmetrical Liang Bu weeding mechanism, between Liang Bu weeding mechanism, inside and outside travel mechanism is set, described inside and outside travel mechanism comprises flat-removing air cyclinder, the expansion link of described flat-removing air cyclinder connects the weeding mechanism being positioned at outside, described weeding mechanism comprises upper comb dent plate and lower fishback, between described upper comb dent plate and lower fishback, cutter is set, the front end of described upper comb dent plate and lower fishback arranges teeth in a row, the front side edge edge of described cutter is blade, described lower fishback arranges device for administration of drugs, described integrated manipulator electricly connects flat-removing air cyclinder and device for administration of drugs.
2. intelligent weeder according to claim 1, is characterized in that, the blade indention shape of described cutter, and the sawtooth wedge angle of described zigzag fashion is right angle or acute angle.
3. intelligent weeder according to claim 1, is characterized in that, the inner side of described lower fishback sets firmly ledger-plate, and described cutter connects movable mechanism.
4. intelligent weeder according to claim 3, it is characterized in that, described movable mechanism comprises electromotor, the rotating shaft of described electromotor drives connecting eccentric wheel, described eccentric wheel is connected with connecting rod, the driving end of described connecting rod is hinged on above-mentioned cutter, and corresponding the formation with upper comb dent plate or lower fishback in the two sides up and down of described cutter is slidably connected, and described integrated manipulator electricly connects electromotor.
5. intelligent weeder according to claim 1, is characterized in that, the teeth of described upper comb dent plate are downward-sloping, and its angle of inclination is 35 ° to 75 °, and the increment of described upper comb dent plate teeth is gathered close with the increment of lower fishback teeth mutually.
6. intelligent weeder according to claim 1, it is characterized in that, described device for administration of drugs comprises deposits medicine-chest, described medicine-chest of depositing connects grug transportation tube, the import of described grug transportation tube arranges electrically-controlled valve, the end of described grug transportation tube connects medicine dropper, the teeth bottom side of described lower fishback arranges dispenser sponge, the mouth of pipe of described medicine dropper contacts above-mentioned dispenser sponge, humidity sensor is placed in described dispenser sponge, described humidity sensor connects above-mentioned integrated manipulator by circuit, and described integrated manipulator is electrically connected above-mentioned electrically-controlled valve.
7. intelligent weeder according to claim 1, it is characterized in that, the bottom side of described workbench arranges fore-stock and after-poppet, the bottom of described fore-stock is fixedly mounted with the weeding mechanism of above-mentioned weeder in the ranks, described after-poppet has two sub-stents, and the bottom of described sub-stent is fixedly mounted with the weeding mechanism of weeder in above-mentioned row, and the bottom surface of described workbench sets firmly guide rail, the top of described sub-stent arranges roller, and described roller embeds in guide rail to be formed and is slidably connected.
8. intelligent weeder according to claim 1, it is characterized in that, described travelling car also comprises four wheels, described four wheels are all connected to below workbench by double cross arm independent suspension, double cross arm independent suspension comprises Top Crossbeam and lower cross arm, and Top Crossbeam is hinged on the bottom surface of workbench, and vehicle wheel rotation is connected on lower cross arm, be hinged between Top Crossbeam and lower cross arm, between Top Crossbeam and lower cross arm, damper be housed.
9. a weeds sweep-out method for intelligent weeder, adopts intelligent weeder as claimed in claim 1, it is characterized in that, comprise the following steps:
1), by intelligent weeder be positioned over one end in the ranks, make the weeder in the ranks on front side of it be positioned in the ranks, in the row of rear side, weeder stretches in the adjacent inline crops in both sides;
2), intelligent weeder walks forward, is positioned at weeding mechanism in the ranks directly by upper comb dent plate and lower fishback combing weeds, then cuts off weeds by cutter, finally by device for administration of drugs to the weeds otch spread application just cut;
3), by the scene in the adjacent lines of camera shooting both sides, scene information is passed to integrated manipulator analysis, and analytic process and result is passed through Computer display out, comprise the method for following several identification crops:
I, color characteristic identification is utilized:
Point by point scanning is carried out to the image collected, splits green planting and soil according to the rgb value of each pixel, namely utilize color characteristic first from Soil Background, to isolate green planting, and then distinguish weeds and crop by the color characteristic detecting target area;
II, morphological feature identification is utilized:
Plants identification based on morphological feature realizes mainly through the border style and whole blade shape studying blade; The blade utilizing deformable template method to carry out identification division to block: the end first detecting blade, then model is placed on each end of blade, and by distortion, model being mated with whole blade, model structure is determined with the outline line representing blade shape by the variable skeleton of blade;
III, spectral signature identification is utilized:
Under the irradiation of visible ray or near infrared light, the spectral reflectivity of weeds and soil and weeds and crop also exists difference;
IV, textural characteristics identification is utilized:
Whether texture is the repeated arrangement of certain basic model, can occur significantly to change the border determined between two kinds of texture pattern by observing texture measure;
V, pattern-recongnition method:
Artificial neural network, SVMs, Bayes classifier, decision tree and discriminant analysis pattern is utilized to carry out identification weeds and crop;
4), identify and greenly plant for crop, integrated manipulator manipulation flat-removing air cyclinder retraction expansion link, drive weeding mechanism to move to the inside to realize dodging to leave inline crops, intelligence weeder continues to move ahead, until camera detection is completely through crop, integrated manipulator manipulates flat-removing air cyclinder again and stretches out expansion link, drives the movement laterally of weeding mechanism to reenter space in row;
5), identify green planting for weeds, stretch into row in weeding mechanism by upper comb dent plate and lower fishback combing weeds, then cut off weeds by cutter, finally by device for administration of drugs to the weeds otch spread application just cut.
10. weeds sweep-out method according to claim 9, it is characterized in that, the humidity information of described humidity sensor sensing dispenser sponge, and humidity information is sent to integrated manipulator in real time, integrated manipulator carries out the judgement of dispenser sponge herb liquid content by detecting data, once judge that liquid is containing quantity not sufficient, integrated manipulator controls electrically-controlled valve and opens, making liquid by depositing medicine-chest through grug transportation tube, dropping on dispenser sponge by medicine dropper; Until the sensing humidity value of humidity sensor is up to standard, integrated manipulator carries out dose judgement by detecting data, controls electrically-controlled valve further and closes, complete feeding operations.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510746468.XA CN105230224B (en) | 2015-11-06 | 2015-11-06 | Intelligent weeding machine and weed removing method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510746468.XA CN105230224B (en) | 2015-11-06 | 2015-11-06 | Intelligent weeding machine and weed removing method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105230224A true CN105230224A (en) | 2016-01-13 |
| CN105230224B CN105230224B (en) | 2017-03-22 |
Family
ID=55028296
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510746468.XA Active CN105230224B (en) | 2015-11-06 | 2015-11-06 | Intelligent weeding machine and weed removing method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105230224B (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105961366A (en) * | 2016-06-15 | 2016-09-28 | 邢永安 | Garden weeding robot |
| CN106471997A (en) * | 2017-01-03 | 2017-03-08 | 朱轩昂 | A kind of medicine device smeared based on hay mover |
| CN106718388A (en) * | 2016-11-22 | 2017-05-31 | 邱逸奎 | Quickly embrocate herbicidal methods |
| WO2017215624A1 (en) * | 2016-06-16 | 2017-12-21 | 苏州科瓴精密机械科技有限公司 | Robot |
| CN107593200A (en) * | 2017-10-31 | 2018-01-19 | 河北工业大学 | A kind of trees plant protection system and method based on visible ray infrared technique |
| CN107926266A (en) * | 2017-12-07 | 2018-04-20 | 枞阳县共义生态农业科技有限公司 | A kind of rideable Novel grass cutter of fruit-bearing forest |
| CN108012592A (en) * | 2017-12-08 | 2018-05-11 | 吉林大学 | Seedling weeder is kept away in a kind of intertillage phase corn field |
| CN108450122A (en) * | 2017-12-15 | 2018-08-28 | 雷可强 | A kind of efficient grass trimmer |
| CN108812622A (en) * | 2018-07-11 | 2018-11-16 | 安徽合众智信科技咨询有限公司 | A kind of multiduty agriculture farmland weeding is mechanical |
| CN108901333A (en) * | 2018-07-27 | 2018-11-30 | 徐亮 | A kind of antiwind lawn trimming attachment in novel gardens |
| CN108982420A (en) * | 2018-07-12 | 2018-12-11 | 华南农业大学 | A kind of the rice crop row Midline positioning device and method of light-tactilely-perceptible |
| CN109601029A (en) * | 2018-12-25 | 2019-04-12 | 金华市众鑫农业科技有限公司 | A kind of agricultural weeding device |
| CN111837593A (en) * | 2020-07-24 | 2020-10-30 | 安徽农业大学 | Novel peucedanum praeruptorum weeding machine based on machine vision and convolutional neural network algorithm |
| CN112335351A (en) * | 2020-11-06 | 2021-02-09 | 太原理工大学 | Lever puller with image recognition function |
| CN113349188A (en) * | 2021-05-31 | 2021-09-07 | 南京林业大学 | Lawn and forage grass precise weeding method based on cloud weeding spectrum |
| CN114080889A (en) * | 2021-11-17 | 2022-02-25 | 东北农业大学 | Rice leaf age index management digital diagnosis unmanned vehicle |
| CN114489045A (en) * | 2021-12-27 | 2022-05-13 | 广州极飞科技股份有限公司 | Operation control method, module, equipment and storage medium |
| CN114651596A (en) * | 2022-05-25 | 2022-06-24 | 苏州农业职业技术学院 | Intelligent leveling device of unmanned mower and implementation method |
| CN115067056A (en) * | 2022-06-13 | 2022-09-20 | 扬州大学 | Horizontal control method and system for weeding machine |
| CN115088472A (en) * | 2022-06-21 | 2022-09-23 | 河南农业大学 | Wound type weeding robot based on machine vision |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201054889Y (en) * | 2007-01-25 | 2008-05-07 | 上海罗宝信息技术有限公司 | Remote-network solar automatic mowing machine |
| US20080245043A1 (en) * | 2007-04-05 | 2008-10-09 | Shinn Rickey D | Rotary cutter device with integrated cooling system |
| CN202077380U (en) * | 2011-04-01 | 2011-12-21 | 上海大学 | Global positioning system (GPS) navigation mowing robot |
| CN203072359U (en) * | 2013-01-28 | 2013-07-24 | 徐艳东 | Front-located folding type wide-width Stipa baicalensis ear removing machine |
| US20130238130A1 (en) * | 2012-03-06 | 2013-09-12 | Travis Dorschel | Path recording and navigation |
| US20150085126A1 (en) * | 2011-11-17 | 2015-03-26 | Tzvi Avnery | Lawn Mower |
| CN205082548U (en) * | 2015-11-06 | 2016-03-16 | 嘉兴职业技术学院 | Intelligence weeder |
-
2015
- 2015-11-06 CN CN201510746468.XA patent/CN105230224B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201054889Y (en) * | 2007-01-25 | 2008-05-07 | 上海罗宝信息技术有限公司 | Remote-network solar automatic mowing machine |
| US20080245043A1 (en) * | 2007-04-05 | 2008-10-09 | Shinn Rickey D | Rotary cutter device with integrated cooling system |
| CN202077380U (en) * | 2011-04-01 | 2011-12-21 | 上海大学 | Global positioning system (GPS) navigation mowing robot |
| US20150085126A1 (en) * | 2011-11-17 | 2015-03-26 | Tzvi Avnery | Lawn Mower |
| US20130238130A1 (en) * | 2012-03-06 | 2013-09-12 | Travis Dorschel | Path recording and navigation |
| CN203072359U (en) * | 2013-01-28 | 2013-07-24 | 徐艳东 | Front-located folding type wide-width Stipa baicalensis ear removing machine |
| CN205082548U (en) * | 2015-11-06 | 2016-03-16 | 嘉兴职业技术学院 | Intelligence weeder |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105961366B (en) * | 2016-06-15 | 2018-11-20 | 重庆至和园林开发有限公司 | A kind of gardens weed-eradicating robot |
| CN105961366A (en) * | 2016-06-15 | 2016-09-28 | 邢永安 | Garden weeding robot |
| WO2017215624A1 (en) * | 2016-06-16 | 2017-12-21 | 苏州科瓴精密机械科技有限公司 | Robot |
| CN106718388A (en) * | 2016-11-22 | 2017-05-31 | 邱逸奎 | Quickly embrocate herbicidal methods |
| CN106471997A (en) * | 2017-01-03 | 2017-03-08 | 朱轩昂 | A kind of medicine device smeared based on hay mover |
| CN107593200A (en) * | 2017-10-31 | 2018-01-19 | 河北工业大学 | A kind of trees plant protection system and method based on visible ray infrared technique |
| CN107593200B (en) * | 2017-10-31 | 2022-05-27 | 河北工业大学 | Tree plant protection system and method based on visible light-infrared technology |
| CN107926266A (en) * | 2017-12-07 | 2018-04-20 | 枞阳县共义生态农业科技有限公司 | A kind of rideable Novel grass cutter of fruit-bearing forest |
| CN108012592A (en) * | 2017-12-08 | 2018-05-11 | 吉林大学 | Seedling weeder is kept away in a kind of intertillage phase corn field |
| CN108450122A (en) * | 2017-12-15 | 2018-08-28 | 雷可强 | A kind of efficient grass trimmer |
| CN108812622A (en) * | 2018-07-11 | 2018-11-16 | 安徽合众智信科技咨询有限公司 | A kind of multiduty agriculture farmland weeding is mechanical |
| CN108812622B (en) * | 2018-07-11 | 2020-12-22 | 重庆市潼南区太安农业综合开发有限公司 | Agricultural farmland weeding machine of multipurpose |
| CN108982420A (en) * | 2018-07-12 | 2018-12-11 | 华南农业大学 | A kind of the rice crop row Midline positioning device and method of light-tactilely-perceptible |
| CN108982420B (en) * | 2018-07-12 | 2020-01-17 | 华南农业大学 | Light-touch perception rice crop row center line positioning device and method |
| CN108901333B (en) * | 2018-07-27 | 2021-08-31 | 福建省佳志成建设工程有限公司 | Gardens are with antiwind lawn trimming means |
| CN108901333A (en) * | 2018-07-27 | 2018-11-30 | 徐亮 | A kind of antiwind lawn trimming attachment in novel gardens |
| CN109601029A (en) * | 2018-12-25 | 2019-04-12 | 金华市众鑫农业科技有限公司 | A kind of agricultural weeding device |
| CN111837593A (en) * | 2020-07-24 | 2020-10-30 | 安徽农业大学 | Novel peucedanum praeruptorum weeding machine based on machine vision and convolutional neural network algorithm |
| CN112335351A (en) * | 2020-11-06 | 2021-02-09 | 太原理工大学 | Lever puller with image recognition function |
| CN113349188A (en) * | 2021-05-31 | 2021-09-07 | 南京林业大学 | Lawn and forage grass precise weeding method based on cloud weeding spectrum |
| CN113349188B (en) * | 2021-05-31 | 2022-06-03 | 南京林业大学 | Lawn and forage precise weeding method based on cloud weeding spectrum |
| CN114080889A (en) * | 2021-11-17 | 2022-02-25 | 东北农业大学 | Rice leaf age index management digital diagnosis unmanned vehicle |
| CN114489045A (en) * | 2021-12-27 | 2022-05-13 | 广州极飞科技股份有限公司 | Operation control method, module, equipment and storage medium |
| CN114651596A (en) * | 2022-05-25 | 2022-06-24 | 苏州农业职业技术学院 | Intelligent leveling device of unmanned mower and implementation method |
| CN114651596B (en) * | 2022-05-25 | 2022-09-02 | 苏州农业职业技术学院 | Intelligent leveling device of unmanned mower and implementation method |
| CN115067056A (en) * | 2022-06-13 | 2022-09-20 | 扬州大学 | Horizontal control method and system for weeding machine |
| CN115067056B (en) * | 2022-06-13 | 2023-12-08 | 扬州大学 | Horizontal control method and system for weeding machine |
| CN115088472A (en) * | 2022-06-21 | 2022-09-23 | 河南农业大学 | Wound type weeding robot based on machine vision |
| CN115088472B (en) * | 2022-06-21 | 2024-02-02 | 河南农业大学 | Wound weeding robot based on machine vision |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105230224B (en) | 2017-03-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105230224B (en) | Intelligent weeding machine and weed removing method thereof | |
| Steward et al. | The use of agricultural robots in weed management and control | |
| US10568316B2 (en) | Apparatus and methods for in-field data collection and sampling | |
| Esau et al. | Machine vision smart sprayer for spot-application of agrochemical in wild blueberry fields | |
| CN102564593B (en) | Plant growth condition monitoring system based on compute vision and internet of things | |
| CN110163138B (en) | Method for measuring and calculating wheat tillering density based on multispectral remote sensing image of unmanned aerial vehicle | |
| Nieuwenhuizen et al. | Performance evaluation of an automated detection and control system for volunteer potatoes in sugar beet fields | |
| CN104521936B (en) | Weeds clear up system automatically | |
| Ma et al. | Automatic detection of crop root rows in paddy fields based on straight-line clustering algorithm and supervised learning method | |
| CN102239756A (en) | Intelligent farming robot in greenhouse | |
| Vikram | Agricultural Robot–A pesticide spraying device | |
| CN205082548U (en) | Intelligence weeder | |
| CN113724210A (en) | Crop growth recognition method and system | |
| CN117496356A (en) | Agricultural artificial intelligent crop detection method and system | |
| Xu et al. | Research and application on corn crop identification and positioning method based on Machine vision | |
| CN205082547U (en) | Inrow weeding device of intelligence weeder | |
| Jiang et al. | A conceptual evaluation of a weed control method with post-damage application of herbicides: A composite intelligent intra-row weeding robot | |
| Mahmud et al. | Measuring tree canopy density using a lidar-guided system for precision spraying | |
| CN202111998U (en) | Intelligent tillage robot in greenhouse | |
| CN104542551B (en) | A kind of weeds automatic cleaning method | |
| Moghaddam et al. | Developing a selective thinning algorithm in sugar beet fields using machine vision system | |
| Lakshmi et al. | A Review on Developing Tech-Agriculture using Deep Learning Methods by Applying UAVs | |
| CN107389594B (en) | Terahertz light-based main stem terminal bud identification and positioning method and device | |
| US11763452B2 (en) | Training method, evaluation method, electronic device and storage medium | |
| Valero et al. | Single Plant Fertilization using a Robotic Platform in an Organic Cropping Environment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |