CN110274557A - A kind of blade area measuring device based on computer vision and method - Google Patents
A kind of blade area measuring device based on computer vision and method Download PDFInfo
- Publication number
- CN110274557A CN110274557A CN201910522389.9A CN201910522389A CN110274557A CN 110274557 A CN110274557 A CN 110274557A CN 201910522389 A CN201910522389 A CN 201910522389A CN 110274557 A CN110274557 A CN 110274557A
- Authority
- CN
- China
- Prior art keywords
- blade
- measurement
- servo electric
- magnet ring
- cabinet
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/28—Measuring arrangements characterised by the use of optical techniques for measuring areas
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a kind of blade area measuring device based on computer vision and methods, more particularly to field of machine vision, including workbench, the bench-top is equipped with measurement body, the first servo electric jar is equipped at the top of the measurement body, the quantity of first servo electric jar is set as two, the output shaft top of two first servo electric jars is connected with mandril, the mandril bottom is equipped with push rod, the push rod two sides are equipped with first connecting rod, the first connecting rod bottom end is equipped with connecting tube, the push rod bottom end is equipped with measurement case, the measurement internal body is equipped with collection of images device；The measurement case includes cabinet, and the top of the box is equipped with position-limiting tube, is embedded with the first magnet ring on the cabinet exterior.The simple precision of operation of the present invention is high, and effective solution traditional measurement method is big to blade damage, is also easy to produce the problem of thinking error, can accurately measure the slight change of blade area, obtains high-precision blade area information.
Description
Technical field
The present invention relates to field of machine vision, it is more particularly related to a kind of blade based on computer vision
Area measuring device and method.
Background technique
Blade is the important biorgan of crop, and photosynthesis, respiration, transpiration are executed by blade organ
, the variation of blade area be able to reflect leaf photosynthesis product amounts accumulation number, be able to reflect crop plant to light warm water
The case where nutrien utilizations such as fertilizer.Therefore can quickly, accurate, easy, lossless acquisition blade area, to the high yield for formulating crop
Efficient Cultivation strategy plays facilitation.
Traditional method for measuring leaf area includes empirical formula method, nine official's lattice methods, weight method, and empirical formula method passes through leaf
Piece length and width and empirical coefficient estimate area value, and precision is not high.Nine official's lattice methods are by blade tiling to being decorated with fixed size grid
Plate on, calculate blade covering grid number, for the judgement subjectivity for the grid area that blade edge cannot be completely covered
It is too strong, measurement result poor repeatability.Weight method is the leaf area that a known area is taken on blade, by weight ratio after weighing
Transformed area value, this method have destructiveness, can not persistently carry out on same blade.
Summary of the invention
In order to overcome the drawbacks described above of the prior art, the embodiment of the present invention provides a kind of blade based on computer vision
Area measuring device and method are input to computer-internal by the feux rouges image picture for obtaining multiple camera units, and divide
It is made into multiple groups image, multiple groups image is made to form training set, every group of image is collectively labeled as a classification, is carried out using the training set
Surface identification scans input training set using selective search algorithm in R-CNN, finds possible object therein, raw
Multiple regions suggestion is produced, then a convolutional neural networks are run on these regions are suggested, finally, by each convolutional Neural net
Network is exported to SVM, using the bounding box of a linear regression tightening object, is read comprising object object edges frame in training set
It takes, the area of blade is calculated using object frame, precision easy to operate is high, and effective solution traditional measurement method is broken to blade
Bad property is big, is also easy to produce the problem of thinking error, can accurately measure the slight change of blade area.
To achieve the above object, the invention provides the following technical scheme: a kind of blade area based on computer vision is surveyed
Device and method, including workbench are measured, the bench-top is equipped with measurement body, is equipped with first at the top of the measurement body and watches
Electric cylinder is taken, the quantity of first servo electric jar is set as two, the output shaft top of two first servo electric jars
End is connected with mandril, and the mandril bottom is equipped with push rod, and the push rod two sides are equipped with first connecting rod, the first connecting rod bottom
End is equipped with connecting tube, and the push rod bottom end is equipped with measurement case, and the measurement internal body is equipped with collection of images device；
The measurement case includes cabinet, and the top of the box is equipped with position-limiting tube, is embedded with the first magnet ring, institute on the cabinet exterior
Cabinets cavity is stated equipped with pressing block, the pressing block top is equipped with connecting spring, and the pressing block bottom end offers light room, institute
It states light chamber interior and is equipped with multiple red lamp beads, light room bottom is equipped with the first acrylic board, opens on the cabinet bottom wall
Equipped with multiple connecting holes, the connecting hole inner cavity top is equipped with electromagnet, and the bottom of box, which is equipped with, places cover board, the placement
It is equipped with multiple magnetic bars at the marginal surface of cover board, offers placing groove in the middle part of the placement cover board, the placement trench bottom is equipped with
Second acrylic board；
The collection of images device includes seal pipe, and the seal pipe outer cover is equipped with the second magnet ring, and the seal pipe bottom end is set
There is image platform, collecting chamber is offered at the top of the image platform, the collecting chamber intracavity bottom is equipped with multiple camera units, the phase
The top of machine unit is equipped with third acrylic board, and image platform two sides are equipped with the second servo electric jar, second servo
The output shaft top of electric cylinder is equipped with the second connecting rod.
In a preferred embodiment, the workbench bottom is equipped with support rod, and the support rod bottom end, which is equipped with, lives
Driving wheel.
In a preferred embodiment, the measurement case passes through push rod, mandril and the first servo electric jar and measurement
Body activities connection, the first connecting rod top are fixedly connected with mandril, and the connecting tube bottom end is fixedly installed on measurement case
Top, the first connecting rod is flexibly connected with connecting tube.
In a preferred embodiment, the position-limiting tube runs through the top of cabinet, and the push rod bottom end is through limit
It is fixedly connected, is fixedly connected at the top of the connecting spring top and cabinets cavity, outside first magnet ring at the top of pipe and pressing block
Wall is coplanar with cabinet exterior, and the pressing block is flexibly connected by connecting spring with cabinet.
In a preferred embodiment, the seal pipe is embedded at the top of measurement body, the seal pipe top
It is coplanar with at the top of measurement body, the third acrylic board is embedded at the top of image platform, and the seal pipe is set to the
The surface of three acrylic boards.
In a preferred embodiment, the fixed company in second servo electric jar bottom end and measuring machine intracoelomic cavity bottom
It connects, the second connection boom end is fixedly connected with the second magnet ring, and second magnet ring passes through the second connecting rod and the second servo
Electric cylinder is flexibly connected with sealing pipe outer wall, and the measurement body side is equipped with console.
A kind of method for measuring leaf area based on computer vision, specifically comprises the following steps
Step 1 sends control command using console, and control electromagnet power-off will be placed cover board and be taken out from measurement bottom portion,
The blade won is placed on the second acrylic board in placing groove, is paved, electromagnet is then controlled and is powered, frid will be placed
The magnetic bar alignment connecting hole at top is inserted into, and magnetic bar and electromagnet are adsorbed, and will be placed frid and is fixed on measurement bottom portion；
Step 2 sets the protrusion height of the second servo-electric output shaft, the height of the second magnet ring is fixed, control first
Servo-electric cylinder working sets the moving distance of the first servo electric jar according to the distance between the first magnet ring and the second magnet ring,
First servo electric jar drives mandril mobile, and the measurement case of push rod bottom end is moved to inside seal pipe, when the first magnet ring and second
When magnet ring adsorbs, the first servo electric jar stops working, and then controls the first servo electric jar and continues to push down on 1cm, makes to push away
Bar squeezes pressing block in gag lever post, and the first acrylic board that pressing block moves down its bottom is pressed together on the leaf on the second acrylic board
Piece surface, shakeouts blade in placing groove；
Step 3 is controlled multiple red lamp bead work, is irradiated using the feux rouges of long wavelength to blade, is not blocked by blade
Feux rouges is incident upon in collecting chamber by the second acrylic board and third acrylic board, controls multiple camera unit work at this time, right
Blade image carries out capture shooting；
The feux rouges image picture that multiple camera units obtain is input to computer-internal, and is distributed into multiple groups image by step 4,
Multiple groups image is set to form training set, every group of image is collectively labeled as a classification, carries out surface identification using the training set；
Step 5, the convolutional neural networks in image processing unit are obtained training set and are calculated in R-CNN using selective search
Method scanning input training set, finds possible object therein, produces multiple regions suggestion, then run on these regions are suggested
One convolutional neural networks tightens object using a linear regression finally, exporting each convolutional neural networks to SVM
Bounding box is read out in training set comprising object object edges frame；
Step 6 calculates the area of each object frame using Heron's formula, so according to the length and width data of the object frame of acquisition
The area of measurement blade is obtained after being superimposed afterwards.
Technical effect and advantage of the invention:
By setting the protrusion height of the second servo-electric output shaft, the height of the second magnet ring is fixed, control first is watched
Electronic cylinder working is taken, according to the distance between the first magnet ring and the second magnet ring, sets the moving distance of the first servo electric jar, the
One servo electric jar drives mandril mobile, and the measurement case of push rod bottom end is moved to inside seal pipe, when the first magnet ring and the second magnetic
When ring adsorbs, the first servo electric jar stops working, and then controls the first servo electric jar and continues to push down on 1cm, makes push rod
Pressing block is squeezed in gag lever post, the first acrylic board that pressing block moves down its bottom is pressed together on the blade on the second acrylic board
Surface shakeouts blade in placing groove, controls multiple red lamp bead work, is shone using the feux rouges of long wavelength blade
It penetrates, is not incident upon in collecting chamber by the feux rouges that blade blocks by the second acrylic board and third acrylic board, controlled at this time more
A camera unit work carries out blade image to capture the actual profile error for shooting the image information and blade that make to obtain most
Small, the image information of acquisition is truer, when being calculated using convolutional neural networks blade area, obtains high-precision
Blade area information；
It is input to computer-internal by the feux rouges image picture for obtaining multiple camera units, and is distributed into multiple groups image, is made
Multiple groups image forms training set, and every group of image is collectively labeled as a classification, carries out surface identification using the training set,
In R-CNN, input training set is scanned using selective search algorithm, possible object therein is found, produces multiple regions suggestion,
Then one convolutional neural networks of operation make finally, exporting each convolutional neural networks to SVM on these regions are suggested
Object edges are utilized to including that object object edges frame is read out in training set with the bounding box of a linear regression tightening object
Frame calculates the area of blade, precision easy to operate is high, and effective solution traditional measurement method is big to blade damage, is also easy to produce
The problem of thinking error can accurately measure the slight change of blade area.
Detailed description of the invention
Fig. 1 is overall structure diagram one of the invention.
Fig. 2 is structural schematic diagram when measurement case of the invention moves down.
Fig. 3 is overall structure diagram two of the invention.
Fig. 4 is measurement box structure schematic diagram of the invention.
Fig. 5 is detailed structure schematic diagram at A in Fig. 2 of the invention.
Fig. 6 is placement covering plate structure schematic diagram of the invention.
Appended drawing reference are as follows: 1 workbench, 2 measurement bodies, 3 first servo electric jars, 4 mandrils, 5 push rods, 6 first connecting rods,
7 connecting tubes, 8 measurement casees, 9 collection of images devices, 10 cabinets, 11 position-limiting tubes, 12 first magnet rings, 13 pressing blocks, 14 connecting springs,
Cover board, 21 magnetic bars, 22 placements are placed in 15 light rooms, 16 red lamp beads, 17 first acrylic boards, 18 connecting holes, 19 electromagnet, 20
Sub- gram of slot, 23 second acrylic boards, 24 seal pipes, 25 second magnet rings, 26 image platforms, 27 collecting chambers, 28 camera units, 29 thirds
Power plate, 30 second servo electric jars, 31 second connecting rods.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
A kind of blade area measuring device based on computer vision and method as shown in figures 1 to 6, including workbench 1,
It is equipped with measurement body 2 at the top of the workbench 1, is equipped with the first servo electric jar 3 at the top of the measurement body 2, described first watches
The quantity for taking electric cylinder 3 is set as two, and the output shaft top of two first servo electric jars 3 is connected with mandril 4, described
4 bottom of mandril is equipped with push rod 5, and 5 two sides of push rod are equipped with first connecting rod 6, and 6 bottom end of first connecting rod is equipped with connecting tube
7,5 bottom end of push rod is equipped with measurement case 8, is equipped with collection of images device 9 inside the measurement body 2；
The measurement case 8 includes cabinet 10, and position-limiting tube 11 is equipped at the top of the cabinet 10, and the is embedded on 10 outer wall of cabinet
One magnet ring 12,10 inner cavity of cabinet are equipped with pressing block 13, are equipped with connecting spring 14, the pressing block at the top of the pressing block 13
13 bottom ends offer light room 15, multiple red lamp beads 16 are equipped with inside the light room 15,15 bottom of light room is equipped with
First acrylic board 17, multiple connecting holes 18 are offered on 10 bottom wall of cabinet, and 18 inner cavity top of connecting hole is equipped with electricity
Magnet 19,10 bottom of cabinet, which is equipped with, places cover board 20, is equipped with multiple magnetic bars 21 at the marginal surface for placing cover board 20,
Placing groove 22 is offered in the middle part of the placement cover board 20,22 bottom of placing groove is equipped with the second acrylic board 23；
The collection of images device 9 includes seal pipe 24, and 24 outer cover of seal pipe is equipped with the second magnet ring 25, the seal pipe
24 bottom ends are equipped with image platform 26, and collecting chamber 27 is offered at the top of the image platform 26, and 27 intracavity bottom of collecting chamber is equipped with more
A camera unit 28, the top of the camera unit 28 are equipped with third acrylic board 29, and 26 two sides of image platform are equipped with the
The output shaft top of two servo electric jars 30, second servo electric jar 30 is equipped with the second connecting rod 31.
1 bottom of workbench is equipped with support rod, and the support rod bottom end is equipped with castor, and the measurement case 8 is by pushing away
Bar 5, mandril 4 and the first servo electric jar 3 are flexibly connected with measurement body 2, and 6 top of first connecting rod and mandril 4 are fixed
Connection, 7 bottom end of connecting tube are fixedly installed on the top of measurement case 8, and the first connecting rod 6 is flexibly connected with connecting tube 7；
The position-limiting tube 11 runs through the top of cabinet 10, and 5 bottom end of push rod is fixed at the top of position-limiting tube 11 and pressing block 13
Connection, 14 top of connecting spring are fixedly connected with 10 inner cavity top of cabinet, outside 12 outer wall of the first magnet ring and cabinet 10
Wall is coplanar with, and the pressing block 13 is flexibly connected by connecting spring 14 with cabinet 10；
The seal pipe 24 is embedded at the top of measurement body 2, and 24 top of seal pipe is coplanar with 2 top of measurement body to be set
It sets, the third acrylic board 29 is embedded at the top of image platform 26, and the seal pipe 24 is set to third acrylic board 29
Surface；
Second servo electric jar, 30 bottom end is fixedly connected with measurement 2 intracavity bottom of body, 31 end of the second connecting rod
It is fixedly connected with the second magnet ring 25, second magnet ring 25 passes through the second connecting rod 31 and the second servo electric jar 30 and seal pipe
24 outer walls are flexibly connected, and 2 side of measurement body is equipped with console.
Embodiment specifically: using console send control command, control electromagnet 19 power off, will place cover board 20 from
It measures 8 bottom of case to take out, the blade won is placed on the second acrylic board 23 in placing groove 22, paves, then controls
Electromagnet 19 is powered, and the magnetic bar 21 at the top of 22 plate of placing groove is directed at connecting hole 18 and is inserted into, magnetic bar 21 and electromagnet 19 are made
22 plate of placing groove is fixed on measurement 8 bottom of case, the protrusion height of the second servo-electric output shaft is set, to the second magnetic by absorption
The height of ring 25 is fixed, control the first servo electric jar 3 work, according between the first magnet ring 12 and the second magnet ring 25 away from
From the moving distance of the first servo electric jar 3 of setting, the first servo electric jar 3 drives mandril 4 mobile, the measurement of 5 bottom end of push rod
Case 8 is moved to inside seal pipe 24, and when the first magnet ring 12 and the absorption of the second magnet ring 25, the first servo electric jar 3 stops working,
Then it controls the first servo electric jar 3 to continue to push down on 1cm, push rod 5 is made to squeeze pressing block 13, pressing block 13 in gag lever post
The first acrylic board 17 for moving down its bottom is pressed together on blade surface on the second acrylic board 23, shakeouts blade in placing groove
In 22, controls multiple red lamp beads 16 and work, blade is irradiated using the feux rouges of long wavelength, the feux rouges not blocked by blade
It is incident upon in collecting chamber 27 by the second acrylic board 23 and third acrylic board 29, controls multiple 28 works of camera unit at this time
Make, capture shooting is carried out to blade image, blade area is obtained after handling using computer image, then controls first
Servo electric jar 3 moves up, and measurement case 8 is removed, then dismantled out of seal pipe 24 to cover board 20 is placed, blade is taken out and carries out it
The measurement of his blade.
A kind of method for measuring leaf area based on computer vision, specifically comprises the following steps；
Step 1 sends control command using console, and control electromagnet 19 powers off, and will place cover board 20 from measurement 8 bottom of case
It takes out, the blade won is placed on the second acrylic board 23 in placing groove 22, is paved, it is logical then to control electromagnet 19
Magnetic bar 21 at the top of 22 plate of placing groove is directed at connecting hole 18 and is inserted by electricity, is adsorbed magnetic bar 21 and electromagnet 19, will be placed
22 plate of slot is fixed on measurement 8 bottom of case；
Step 2 sets the protrusion height of the second servo-electric output shaft, the height of the second magnet ring 25 is fixed, control the
The work of one servo electric jar 3 sets the first servo electric jar 3 according to the distance between the first magnet ring 12 and the second magnet ring 25
Moving distance, the first servo electric jar 3 drive mandril 4 mobile, and the measurement case 8 of 5 bottom end of push rod is moved to inside seal pipe 24, when
When first magnet ring 12 and the second magnet ring 25 adsorb, the first servo electric jar 3 stops working, and then controls the first servo electric jar 3
Continue to push down on 1cm, push rod 5 is made to squeeze pressing block 13 in gag lever post, pressing block 13 moves down the first acrylic of its bottom
Plate 17 is pressed together on the blade surface on the second acrylic board 23, shakeouts blade in placing groove 22；
Step 3 is controlled multiple red lamp beads 16 and worked, is irradiated using the feux rouges of long wavelength to blade, do not blocked by blade
Feux rouges be incident upon in collecting chamber 27 by the second acrylic board 23 and third acrylic board 29, control multiple camera units at this time
28 work, carry out capture shooting to blade image；
The feux rouges image picture that multiple camera units 28 obtain is input to computer-internal, and is distributed into multiple groups shadow by step 4
Picture, makes multiple groups image form training set, and every group of image is collectively labeled as a classification, carries out surface knowledge using the training set
Not；
Step 5, the convolutional neural networks in image processing unit are obtained training set and are calculated in R-CNN using selective search
Method scanning input training set, finds possible object therein, produces multiple regions suggestion, then run on these regions are suggested
One convolutional neural networks tightens object using a linear regression finally, exporting each convolutional neural networks to SVM
Bounding box is read out in training set comprising object object edges frame；
Step 6 calculates the area of each object frame using Heron's formula, so according to the length and width data of the object frame of acquisition
The area of measurement blade is obtained after being superimposed afterwards.
Working principle of the present invention:
Referring to Figure of description 1-6, by setting seal pipe 24 and measurement case 8, the blade of measurement is placed in putting in measurement case 8
It sets in slot 22, and is pressed using pressing block 13, when carrying out image capturing, measurement case 8 is placed in seal pipe 24, is utilized
The feux rouges of long wavelength is irradiated blade, and multiple camera units 28 obtain the image information of blade, makes the image information obtained
Minimum with the actual profile error of blade, the image information of acquisition is truer, in utilization convolutional neural networks to blade area
When being calculated, high-precision blade area information is obtained.
The several points that should finally illustrate are: firstly, in the description of the present application, it should be noted that unless otherwise prescribed and
It limits, term " installation ", " connected ", " connection " shall be understood in a broad sense, can be mechanical connection or electrical connection, be also possible to two
Connection inside element, can be directly connected, and "upper", "lower", "left", "right" etc. are only used for indicating relative positional relationship, when
The absolute position for being described object changes, then relative positional relationship may change；
Secondly: the present invention discloses in embodiment attached drawing, relates only to the structure being related to the embodiment of the present disclosure, and other structures can
With reference to being commonly designed, under not conflict situations, the same embodiment of the present invention and different embodiments be can be combined with each other；
Last: the foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, all in essence of the invention
Within mind and principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (7)
1. a kind of blade area measuring device based on computer vision, which is characterized in that including workbench (1), the work
It is equipped with measurement body (2) at the top of platform (1), is equipped with the first servo electric jar (3) at the top of the measurement body (2), described first watches
The quantity for taking electric cylinder (3) is set as two, and the output shaft top of two first servo electric jars (3) is connected with mandril
(4), mandril (4) bottom is equipped with push rod (5), and push rod (5) two sides are equipped with first connecting rod (6), first connection
Bar (6) bottom end is equipped with connecting tube (7), and push rod (5) bottom end is equipped with measurement case (8), is equipped with shadow inside the measurement body (2)
As collection device (9)；
The measurement case (8) includes cabinet (10), is equipped with position-limiting tube (11) at the top of the cabinet (10), cabinet (10) outer wall
On be embedded with the first magnet ring (12), cabinet (10) inner cavity is equipped with pressing block (13), the company of being equipped at the top of the pressing block (13)
It connects spring (14), pressing block (13) bottom end offers light room (15), and multiple red are equipped with inside the light room (15)
Lamp bead (16), light room (15) bottom are equipped with the first acrylic board (17), offer on cabinet (10) bottom wall multiple
Connecting hole (18), connecting hole (18) inner cavity top are equipped with electromagnet (19), and cabinet (10) bottom, which is equipped with, places cover board
(20), multiple magnetic bars (21) are equipped at the marginal surface for placing cover board (20), are offered in the middle part of the placement cover board (20)
Placing groove (22), placing groove (22) bottom are equipped with the second acrylic board (23)；
The collection of images device (9) includes seal pipe (24), and seal pipe (24) outer cover is equipped with the second magnet ring (25), institute
Seal pipe (24) bottom end is stated equipped with image platform (26), image platform (26) top offers collecting chamber (27), the collecting chamber
(27) intracavity bottom is equipped with multiple camera units (28), and the top of the camera unit (28) is equipped with third acrylic board (29),
Image platform (26) two sides are equipped with the second servo electric jar (30), the output shaft top of second servo electric jar (30)
Equipped with the second connecting rod (31).
2. a kind of blade area measuring device based on computer vision according to claim 1, it is characterised in that: described
Workbench (1) bottom is equipped with support rod, and the support rod bottom end is equipped with castor.
3. a kind of blade area measuring device based on computer vision according to claim 1, it is characterised in that: described
Measurement case (8) is flexibly connected by push rod (5), mandril (4) and the first servo electric jar (3) with measurement body (2), and described first
Connecting rod (6) top is fixedly connected with mandril (4), and connecting tube (7) bottom end is fixedly installed on the top of measurement case (8), institute
First connecting rod (6) is stated to be flexibly connected with connecting tube (7).
4. a kind of blade area measuring device based on computer vision according to claim 1, it is characterised in that: described
Position-limiting tube (11) runs through the top of cabinet (10), and push rod (5) bottom end is solid at the top of position-limiting tube (11) and pressing block (13)
Fixed connection, connecting spring (14) top are fixedly connected with cabinet (10) inner cavity top, the first magnet ring (12) outer wall and
Cabinet (10) outer wall is coplanar with, and the pressing block (13) is flexibly connected by connecting spring (14) with cabinet (10).
5. a kind of blade area measuring device based on computer vision according to claim 1, it is characterised in that: described
Seal pipe (24) is embedded at the top of measurement body (2), and seal pipe (24) top is coplanar at the top of measurement body (2) to be set
It sets, the third acrylic board (29) is embedded at the top of image platform (26), and the seal pipe (24) is set to third acrylic
The surface of plate (29).
6. a kind of blade area measuring device based on computer vision according to claim 1, it is characterised in that: described
Second servo electric jar (30) bottom end with measurement body (2) intracavity bottom be fixedly connected, the second connecting rod (31) end with
Second magnet ring (25) is fixedly connected, second magnet ring (25) by the second connecting rod (31) and the second servo electric jar (30) and
Seal pipe (24) outer wall is flexibly connected, and the measurement body (2) side is equipped with console.
7. using a kind of blade area measurement based on computer vision of blade area measuring device described in claim 1-6
Method, which is characterized in that specifically comprise the following steps:
Step 1 sends control command using console, and control electromagnet (19) power-off will place cover board (20) from measurement case
(8) bottom is taken out, and the blade won is placed on the second acrylic board (23) in placing groove (22), paves, then controls
Electromagnet (19) is powered, and magnetic bar (21) alignment connecting hole (18) at the top of placing groove (22) plate is inserted into, magnetic bar (21) are made
It is adsorbed with electromagnet (19), placing groove (22) plate is fixed on measurement case (8) bottom；
Step 2 sets the protrusion height of the second servo-electric output shaft, the height of the second magnet ring (25) is fixed, and controls
First servo electric jar (3) work sets the first servo electricity according to the distance between the first magnet ring (12) and the second magnet ring (25)
The moving distance of dynamic cylinder (3), the first servo electric jar (3) drive mandril (4) mobile, and the measurement case (8) of push rod (5) bottom end is mobile
Internal to seal pipe (24), when the first magnet ring (12) and the second magnet ring (25) are adsorbed, the first servo electric jar (3) stops work
Make, then controls the first servo electric jar (3) and continue to push down on 1cm, push rod (5) is made to squeeze pressing block in gag lever post
(13), the first acrylic board (17) that pressing block (13) moves down its bottom is pressed together on the blade table on the second acrylic board (23)
Face shakeouts blade in placing groove (22)；
Step 3 is controlled multiple red lamp beads (16) and worked, is irradiated using the feux rouges of long wavelength to blade, do not hidden by blade
The feux rouges of gear is incident upon in collecting chamber (27) by the second acrylic board (23) and third acrylic board (29), is controlled at this time multiple
Camera unit (28) work, carries out capture shooting to blade image；
The feux rouges image picture that multiple camera units (28) obtain is input to computer-internal, and is distributed into multiple groups by step 4
Image, makes multiple groups image form training set, and every group of image is collectively labeled as a classification, carries out surface using the training set
Identification；
Step 5, the convolutional neural networks in image processing unit are obtained training set and are calculated in R-CNN using selective search
Method scanning input training set, finds possible object therein, produces multiple regions suggestion, then run on these regions are suggested
One convolutional neural networks tightens object using a linear regression finally, exporting each convolutional neural networks to SVM
Bounding box is read out in training set comprising object object edges frame；
Step 6 calculates the area of each object frame using Heron's formula, so according to the length and width data of the object frame of acquisition
The area of measurement blade is obtained after being superimposed afterwards.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910522389.9A CN110274557B (en) | 2019-06-17 | 2019-06-17 | Blade area measuring device and method based on computer vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910522389.9A CN110274557B (en) | 2019-06-17 | 2019-06-17 | Blade area measuring device and method based on computer vision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110274557A true CN110274557A (en) | 2019-09-24 |
| CN110274557B CN110274557B (en) | 2021-02-09 |
Family
ID=67960911
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910522389.9A Active CN110274557B (en) | 2019-06-17 | 2019-06-17 | Blade area measuring device and method based on computer vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110274557B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114241033A (en) * | 2021-12-17 | 2022-03-25 | 贵州省山地资源研究所 | Ectopic acquisition method for plant leaf area vertical structure |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1952599A (en) * | 2005-10-20 | 2007-04-25 | 中国科学院自动化研究所 | A flat leaf area nondestructive measuring system and method thereof |
| CN101358842A (en) * | 2008-09-19 | 2009-02-04 | 广东省农业科学院水稻研究所 | Plant blade area measuring apparatus |
| JP2011133451A (en) * | 2009-11-27 | 2011-07-07 | Kyushu Univ | Optical vegetation index sensor |
| CN105004284A (en) * | 2015-07-14 | 2015-10-28 | 河南科技大学 | Photoelectric tobacco leaf area measuring device |
| CN205120060U (en) * | 2015-11-09 | 2016-03-30 | 平顶山学院 | Leaf area appearance |
-
2019
- 2019-06-17 CN CN201910522389.9A patent/CN110274557B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1952599A (en) * | 2005-10-20 | 2007-04-25 | 中国科学院自动化研究所 | A flat leaf area nondestructive measuring system and method thereof |
| CN101358842A (en) * | 2008-09-19 | 2009-02-04 | 广东省农业科学院水稻研究所 | Plant blade area measuring apparatus |
| JP2011133451A (en) * | 2009-11-27 | 2011-07-07 | Kyushu Univ | Optical vegetation index sensor |
| CN105004284A (en) * | 2015-07-14 | 2015-10-28 | 河南科技大学 | Photoelectric tobacco leaf area measuring device |
| CN205120060U (en) * | 2015-11-09 | 2016-03-30 | 平顶山学院 | Leaf area appearance |
Non-Patent Citations (1)
| Title |
|---|
| 卫婷婷: "设施黄瓜叶片病害检测仪设计", 《中国优秀硕士学位论文全文数据库 农业科技辑》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114241033A (en) * | 2021-12-17 | 2022-03-25 | 贵州省山地资源研究所 | Ectopic acquisition method for plant leaf area vertical structure |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110274557B (en) | 2021-02-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108387262B (en) | Greenhouse information automatic monitoring method based on suspension type sliding rail platform | |
| CN108376419A (en) | A kind of synthesis Growing state survey method and device of potting romaine lettuce | |
| CN103559708B (en) | Industrial fixed-focus camera parameter calibration device based on side's target model | |
| CN107860366B (en) | Mobile greenhouse crop information measurement and diagnosis system | |
| CN110065074A (en) | A kind of the visual servo laser orientation system and method for picking robot | |
| CN106441442A (en) | Field crop phenotypic information high-pass peer monitoring device and monitoring method | |
| CN110084849A (en) | Has the logistics system of automatic measurement volume and weight | |
| CN103697937A (en) | Environment and plant growth state synergism monitoring and analysis device and method | |
| JP2017148028A (en) | Water content observation device, water content observation method and cultivation device | |
| CN109405755A (en) | A kind of large scale tube sheet aperture and pitch measurement device and measurement method | |
| CN207081412U (en) | A kind of animal body chi ultrasound noncontacting measurement device | |
| US20220053170A1 (en) | Goods information collection device | |
| CN110274557A (en) | A kind of blade area measuring device based on computer vision and method | |
| CN107421466A (en) | A kind of synchronous acquisition device and acquisition method of two and three dimensions image | |
| CN107436340A (en) | A kind of plant root cap Integrated Monitoring System and method | |
| CN206505030U (en) | A kind of automatic line detector of automobile stamping | |
| CN105710045A (en) | Green plant phenotypic and sorting system and method | |
| CN103006233B (en) | Device for detecting curved surface and pressure distribution of foot sole | |
| CN111289513A (en) | Crop root system phenotype acquisition system based on channel | |
| CN207081653U (en) | A kind of planer-type gardening plantlet crop high spectrum image acquisition system | |
| CN207234941U (en) | A kind of digital image acquisition instrument | |
| AU2022211136A1 (en) | A data collection and monitoring system, a controlled environment farming system, devices and related methods | |
| CN109141238A (en) | Bark cambered surface phenotypic data standardized monitoring assistor | |
| CN212205982U (en) | Herbaceous plant location and height and equipment for coverage data acquisition | |
| CN206695759U (en) | Data acquisition device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |