CN108465644A - Artificial intelligence wheat quality inspection robot and quality detecting method - Google Patents
Artificial intelligence wheat quality inspection robot and quality detecting method Download PDFInfo
- Publication number
- CN108465644A CN108465644A CN201810830624.4A CN201810830624A CN108465644A CN 108465644 A CN108465644 A CN 108465644A CN 201810830624 A CN201810830624 A CN 201810830624A CN 108465644 A CN108465644 A CN 108465644A
- Authority
- CN
- China
- Prior art keywords
- wheat
- quality inspection
- sensor
- artificial intelligence
- feeding
- 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
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/02—Measures preceding sorting, e.g. arranging articles in a stream orientating
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/36—Sorting apparatus characterised by the means used for distribution
- B07C5/361—Processing or control devices therefor, e.g. escort memory
Abstract
Artificial intelligence wheat quality inspection robot and quality detecting method, artificial intelligence wheat quality inspection robot include artificial intelligence wheat quality inspection robot body and system server;The artificial intelligence wheat quality inspection robot body includes babinet, the babinet is equipped with feeding mouth and discharge port, feeding mechanism is equipped in the babinet, transport mechanism, upper and lower two-sided machine visual mechanisms and mechanism for sorting, the feeding mouth is connected with feeding mechanism, the feeding mechanism is connected with transport mechanism, the transport mechanism passes through upper and lower two-sided machine visual mechanisms, the transport mechanism is connected with discharge port, the mechanism for sorting is located at upper and lower two-sided machine visual mechanisms rear, the two-sided machine visual mechanisms up and down, mechanism for sorting is connected with system server;The feeding mechanism takes sensor to add the mode of intelligent control.Using the present invention, the detection of wheat unsound grain and impurity can be completed, wheat unsound grain is solved and quickly detects this global problem.
Description
Technical field
The present invention relates to wheat quality physical detection apparatus and method fors, and in particular to a kind of artificial intelligence wheat quality inspection machine
People and quality detecting method.
Background technology
Wheat unsound grain(It is worm, bud, disease, damage, mould)Detection is differentiated by artificial sense completely without modernization means and equipment
Wheat detection is completed, and there are many uncertain factors for artificial sense identification, with portion sample, same inspector is repeatedly reflected
Fixed, the data identified every time all can be different, with portion sample, the same standard of different testing staff, the findings data of identification
Also it can be not quite similar;Accordingly, it is difficult to it is objective, fair, accurately distinguish, come true " have a standard and without the technology of the standard of execution
Means ".
The detection of wheat unsound grain is completed in existing artificial sense identification, is had the following disadvantages:1)Easily by environment and personnel
Factor is interfered;2)Without unified measurement facility;3)Inefficiency.
CN207188252U disclosed a kind of grain unsound grain quality inspection system on April 6th, 2018, was by with spiral
Mechanism feeding, the feeding of optical glass disk, upper and lower double vision feel opposite intake material image, then by detected object and image library
Middle image carries out the identification of mechanical comparison mode, sorts and weigh in glass disk, is finally completed the detection of grain unsound grain.It is this
Detection device has the following problems with means:
(1)Screw mechanism feeding manner is usually applicable only to standard component, when material is the non-standard component of grain class, because of particle
Size, shape, length, proportion, integrated degree difference and be susceptible to putty, material leakage, the discharging situations such as not to the utmost, thus make
The sample that must be conveyed is incomplete and seriously affects efficiency.
(2)Optical glass disk feeding style, when disk rotational reaches certain rotating speed, material is easy to slip, meanwhile, by
It jumps so that the sample deficient in stability in field of vision is sent into, thus further such that the object obtained at the end generated when disk rotational
Severe deviations occur for the raw information of the measurement spatial distortion of material, detected material.
(3)Visual manner opposing upper and lower causes top and bottom visual information to interfere between each other, further causes to be detected
It surveys the distortion of object raw information and generates more large deviation.
(4)The image library of the original image of detected material and master sample progress mechanical comparison is not adapted into grain
Non-standard feature identifies that it is linear to carry out hardness not according to linear relationship using mechanical comparison for grain unsound grain granules
Relationship differentiation misses by a mile in actual operation.
(5)When detected particle reaches sorting area from identified area, physical location is often because of sliding and disk end
It jumps and makes sorting accuracy that can not ensure.
Invention content
The technical problem to be solved by the present invention is in view of the deficiencies of the prior art, provide a kind of to wheat unsound grain
(It is worm, bud, disease, damage, mould)The artificial intelligence wheat quality inspection robot that several groups of five major class and impurity are detected.
The present invention is further the technical problem to be solved is that, offer is a kind of to utilize the artificial intelligence wheat quality inspection machine
The method that people carries out quality inspection to wheat.
The technical solution adopted by the present invention to solve the technical problems is:A kind of artificial intelligence wheat quality inspection robot, packet
Include artificial intelligence wheat quality inspection robot body and system server;The artificial intelligence wheat quality inspection robot body includes case
Body, the babinet are equipped with feeding mouth and discharge port, feeding mechanism, transport mechanism, upper and lower two-sided machine are equipped in the babinet
Visual mechanisms and mechanism for sorting, the feeding mouth are connected with feeding mechanism, and the feeding mechanism is connected with transport mechanism, institute
Stating transport mechanism passes through upper and lower two-sided machine visual mechanisms, the transport mechanism to be connected with discharge port, the mechanism for sorting position
In upper and lower two-sided machine visual mechanisms rear, two-sided machine visual mechanisms, the mechanism for sorting up and down are connected with system server;
The feeding mechanism includes feeding box, and a partition board is equipped in the feeding box, the cavity of feeding box is divided into ante-chamber and back cavity, institute
The hopper that ante-chamber and back cavity are run through in feeding box bottom equipped with one is stated, the back cavity is equipped with monitoring back cavity close to baffle material trough rim is
The no sensor I for having material, the back cavity are equipped with the sensor II of horizontal direction monitoring material position, and the ante-chamber is at partition board
Hopper center is equipped with the sensor III whether monitoring back cavity discharges, and whether ante-chamber hopper exit upper edge is equipped with monitoring ante-chamber
The sensor IV of putty, the feeding box are equipped with the mechanism for preventing back cavity putty, and the feeding box lower part is equipped with vibrator, institute
It states vibrator with voltage modulator to be connected, the sensor I, sensor III, sensor IV, prevents back cavity putty at sensor II
Mechanism, voltage modulator be connected with controller, the controller is connected with system server.
Further, the sensor II of horizontal direction monitoring material position be located at back cavity horizontal direction distance bottom >=1cm
(It is preferred that 2-3 cm)Place.
Further, described to prevent the mechanism of back cavity putty include steering engine, and the steering engine is located at outside feeding box, and the steering engine is logical
It crosses steering engine connecting rod and is connected with steering engine plectrum, the steering engine plectrum is located in back cavity, and close to partition board and hopper.
Further, the transport mechanism includes transparent synchrome conveying belt and synchronous driving wheel, the synchronous driving wheel arrangement
On babinet, the synchronous driving wheel is equipped with synchronous gear I, and the edge of the transparent synchrome conveying belt is equipped with synchronous gear II,
The synchronous gear II of the transparent synchrome conveying belt and I assembly connection of synchronous gear of synchronous driving wheel, the transparent synchrome conveying belt
On partially pass through up and down two-sided machine visual mechanisms.
Further, the two-sided machine visual mechanisms up and down include upper machine vision mechanism, upper shadow-free light source, upper vision bottom
Plate, lower machine vision mechanism, lower shadow-free light source and lower vision bottom plate, the upper machine vision mechanism, upper shadow-free light source are located at saturating
Rear on the upside of bright synchrome conveying belt, the upper vision bottom plate are set on the downside of transparent synchrome conveying belt, and just positioned at upper shadow-free light source
Lower section, the lower machine vision mechanism, lower shadow-free light source are located at front on the downside of transparent conveyor belt, and the lower vision bottom plate is set to saturating
On the upside of bright synchrome conveying belt, and right over lower shadow-free light source.
Further, the mechanism for sorting includes sorting sensor, pulp-collecting box, air tube and weighing sensor, and the sorting passes
Sensor is mounted on pulp-collecting box top, and above transparent synchrome conveying belt, and the air tube passes through pipeline and gas equipped with air valve
Pump is connected, and the sample to be picked that the transparent synchronous driving of face takes, and the weighing sensor is mounted on pulp-collecting box lower part, and realization waits for
Pick weighing for sample.
Further, it is additionally provided with guide item, the guide item is between feeding mechanism and transport mechanism.
A method of quality inspection being carried out to wheat using the artificial intelligence wheat quality inspection robot, is included the following steps:
Wheat measuring samples are reached into feeding mechanism by the feeding mouth of artificial intelligence wheat quality inspection robot body, are passed through simultaneously
Sensor is responsible for detecting measuring samples in feeding mechanism and walks material state, and if there is putty situation, sensor is by putty information
Via controller sends system server to, controls voltage modulator by controller by system server, and then drive feeder
The vibrator of structure vibrates, and measuring samples are sent to transport mechanism for more than by feeding mechanism, wheat is waited for sample by transport mechanism
Product are sent to two-sided machine visual mechanisms up and down, are obtained by two-sided machine visual mechanisms up and down and are tested wheat samples original image,
It is former to every tested wheat samples by the artificial neural network system of the support graphics process built on system server again
Beginning image carries out algorithm process, and the conversion of the features such as the color of every tested wheat samples, texture, profile, size is characterized letter
Breath value inputs artificial neural network system, is compared with the character pair value of information stored on neuron, obtains character pair
The degree of membership of the value of information, then differentiated according to the degree of membership of the character pair value of information with discriminant function, statistic of classification, by dividing
Mechanism is picked by unsound grain and impurity sorting shoulder to shoulder;Finally, the sample that detection is completed is output to discharge port by transport mechanism, it is whole
A quality check process is all automatically performed.
Further, the character pair value of information stored on the neuron obtains in the following manner:It will be sorted in advance
Wheat samples sent to upper and lower Double-plate machine device visual mechanisms through feeding mechanism, transport mechanism, upper and lower two-sided machine visual mechanisms
Every wheat samples original image is obtained, by supporting the artificial neural network system of graphics process original to every wheat samples
Image carries out algorithm process, converts the features such as the color of wheat, texture, profile, size to characteristic information value, and be stored in people
On the neuron of artificial neural networks system, a wheat characteristic information library is consequently formed.
In the present invention, the feeding mechanism takes sensor to add the mode of intelligent control, by the information of sensor senses,
System can decide whether putty, whether empty material, Flow of Goods and Materials whether smooth homogeneous and can be according to these information progress corresponding position
Reason is not in putty, material leakage, the discharging situations such as not to the utmost to ensure that feeding is continuous, stablize, uniformly.
In the present invention, the transparent synchrome conveying belt of the transport mechanism is completed under the driving that synchronous driving wheel passes through motor
The transmission of material, transparent synchrome conveying belt are that middle section is in field of vision location transparency and edge can be complete with synchronous driving wheel
The synchronous belt of engagement.Field of vision is partially transparent to obtain enough clearly images convenient for two-sided machine visual mechanisms up and down, transparent same
Step conveyer belt is driven and uniform motion by motor under stable frequency in synchronous driving wheel, ensures image clearly, detected material
Raw information deviation will not occur.
In the present invention, the two-sided machine visual mechanisms up and down are staggered corresponding position completely so that two vision lens it
Between do not interfere with each other influence obtain original image accuracy and integrality so that do not interfered with each other between the light of two visions
Influence the reliability and stability of visual environment;Ensure that obtaining detected material measures the accurate, true, complete of spatial information with this
It is whole.
In the present invention, the mechanism for sorting is equipped with sensor, can accurately obtain and be sorted material close to the instant of sorting mouth
Accurate location well picks to ensure the accurate of sorting, does not leak and pick.The unsound grain grain that sorting goes out falls into pulp-collecting box, by
Weighing sensor completion is weighed.
Research practice shows using the present invention from wheat measuring samples information is obtained to detection judgement is completed, to each
The detection time of wheat samples is no more than 20 milliseconds, and artificial intelligence detects a 50 grams of standard sample of robot detection(1200
Or so)Only need identification and data statistic analysis can be completed within 3-4 minutes.Whole process does not need human intervention, can effectively solve
Certainly manually the problem of by human factor and external environment influenced that Testing index is caused to drift about of organoleptic detection, meanwhile, robot
Each sample is all detected one by one, does not stay detection blind spot and dead angle.
The actually detected accuracy and repeatability index of the present invention:
Wheat unsound grain detects(By taking 50 grams of every part of sample of national standard as an example)
Whole unsound grains, error≤0.5%, repetitive error≤0.3%;
Injured kernel, error≤0.5%, repetitive error≤0.3%;
Germinate grain, error≤0.5%, repetitive error≤0.3%;
Scab grain(Red mould grain), error≤0.3%, repetitive error≤0.2%;
Scab grain(Black embryo), error≤0.2%, repetitive error≤0.1%;
Broken kernel(It is broken), error≤0.2%, repetitive error≤0.1%;
Moldy kernel, error≤0.3%, repetitive error≤0.2%;
Yellow rice kernel, every 1000 errors are less than 1, repetitive error zero.
Using the present invention, the detection of wheat unsound grain and impurity can be completed, wheat unsound grain is solved and quickly detects this
One global problem.
Description of the drawings
Fig. 1 is the structure diagram of artificial intelligence wheat quality inspection robotic embodiment of the present invention;
Fig. 2 is the appearance diagram of the artificial intelligence wheat quality inspection robot body of embodiment illustrated in fig. 1;
Fig. 3 is the internal structure chart of the artificial intelligence wheat quality inspection robot body of embodiment illustrated in fig. 1;
Fig. 4 is the structural schematic diagram of the feeding mechanism of the artificial intelligence wheat quality inspection robot body of embodiment illustrated in fig. 1;
Fig. 5 is artificial intelligence wheat quality inspection robot controller annexation figure of the present invention;
Fig. 6 is the structural schematic diagram of the transport mechanism of the artificial intelligence wheat quality inspection robot body of embodiment illustrated in fig. 1;
Fig. 7 is the knot of the upper and lower two-sided machine visual mechanisms of the artificial intelligence wheat quality inspection robot body of embodiment illustrated in fig. 1
Structure schematic diagram;
Fig. 8 is the structural schematic diagram of the mechanism for sorting of the artificial intelligence wheat quality inspection robot body of embodiment illustrated in fig. 1;
Fig. 9 is that artificial intelligence wheat quality inspection machine hostage of the present invention examines process flow diagram flow chart.
Specific implementation mode
The invention will be further described with reference to the accompanying drawings and embodiments.
Embodiment
- Fig. 5 referring to Fig.1, a kind of artificial intelligence wheat quality inspection robot, including artificial intelligence wheat quality inspection robot body
1 and system server 2;The artificial intelligence wheat quality inspection robot body 1 includes babinet 1-1, the babinet 1-1 be equipped with into
It is equipped with feeding mechanism 1-4, transport mechanism 1-5, upper and lower two-sided machine vision in material mouth 1-2 and discharge port 1-3, the babinet 1-1
Mechanism 1-6 and mechanism for sorting 1-7, the feeding mouth 1-2 are connected with feeding mechanism 1-4, the feeding mechanism 1-4 and conveyer
Structure 1-5 is connected, and the transport mechanism 1-5 passes through upper and lower two-sided machine visual mechanisms 1-6, the transport mechanism 1-5 and discharging
Mouth 1-3 is connected, and the mechanism for sorting 1-7 is located at the upper and lower two-sided rears machine visual mechanisms 1-6, and the two-sided machine up and down regards
Feel that mechanism 1-6, mechanism for sorting 1-7 are connected with system server 2;
The feeding mechanism 1-4 includes feeding box 1-4-1, a partition board 1-4-2 is equipped in the feeding box 1-4-1, by feeding box
The cavity of 1-4-1 is divided into ante-chamber and back cavity, and the bottoms the feeding box 1-4-1 are equipped with a straight line hopper 1- for running through ante-chamber and back cavity
4-3 is lined up convenient for the wheat grain of irregular grain shape, and the back cavity is equipped with whether monitoring back cavity has material close to baffle material trough rim
I 1-4-4 of sensor, the sensor of horizontal direction monitoring material position is equipped at the 2cm of back cavity horizontal direction distance bottom
II 1-4-5, hopper center of the ante-chamber at partition board is equipped with sensor III 1-4-6 whether monitoring back cavity discharges, described
Ante-chamber hopper exit upper edge is equipped with monitoring ante-chamber, and whether sensor IV 1-4-7, the feeding box 1-4-1 of putty are externally provided with rudder
Machine 1-4-8, the steering engine 1-4-8 are connected with steering engine plectrum 1-4-9, the steering engine plectrum 1-4-9 by steering engine connecting rod and are located at back cavity
It is interior, and close to partition board 1-4-2 and the lower parts straight line hopper 1-4-3, the feeding box 1-4-1 equipped with vibrator 1-4-10, it is described to shake
Dynamic device 1-4-10 is connected with voltage modulator 4, I 1-4-4 of the sensor, II 1-4-5 of sensor, III 1-4-6 of sensor, sensing
IV 1-4-7 of device, steering engine 1-4-8, voltage modulator 4 are connected with controller 3, and the controller 3 is connected with system server 2;
The course of work:
1)II 1-4-5 of sensor of horizontal direction monitoring material position is triggered, and controller 3 obtains information, sends information to system
Server 2, system server 2 will send out stopping sampling instruction, which is used for on-line checking(On-line automatic sampling);
2)The ante-chamber putty of feeding mechanism, whether IV 1-4-7 of sensor of putty is triggered monitoring ante-chamber, and controller 3 obtains the letter
Breath, sends information to system server 2, and system server 2 sends out stopping sampling instruction first, then, system server 2 to
Controller 3 sends out instruction, and controller 3 controls voltage modulator 4, will reinforce vibrator vibration frequency 2 seconds, then restores vibrator
Vibration frequency;If whether the sensor IV 1-4-7 triggerings of putty do not release monitoring ante-chamber, reinforce vibrator vibration frequency again
Rate 2 seconds, 3 times repeatedly, until ante-chamber putty sensor-triggered releases;
3)III 1-4-6 of sensor whether monitoring back cavity discharges is triggered, and whether monitoring back cavity has I 1-4-4 of sensor of material not
It is triggered, illustrates that the back cavity of feeding mechanism putty, controller obtain heat transfer agent, start steering engine 1-4-8, steering engine plectrum 1-
4-9 stirs rear chamber discharge port sample, releases back cavity putty state;After back cavity sample is all conveyed by feeding mechanism, after monitoring
Whether chamber has I 1-4-4 of sensor of material to be triggered, and controller 2 sends out the information that sample all checks out to system server 3.
The input of the controller 3 is connected with each sensor(Access transducing signal-putty transducing signal, full material sensing
Signal), export and be connected with voltage modulator 4, voltage modulator 4 is connected with the vibrator 1-4-10 on feeding mechanism, leads to
The voltage of ovennodulation feeding mechanism upper vibration generator 1-4-10 achievees the purpose that control vibrator 1-4-10 amplitudes.
The system server 2 is the command centre of wheat quality inspection robot, is the brain of wheat quality inspection robot, system
Server is connected by RS232 communication ports with controller, and controller is connected with each sensor, and controller obtains the shape of each sensor
The information of each sensor is sent to system server by state, and system server is sent according to the heat transfer agent of acquisition to controller
Putty processing is completed in instruction;The system server is connected by RJ45 network interfaces with upper and lower two-sided machine visual mechanisms, is obtained
Take pictorial information, denoising then carried out to picture, analyzed, reasoning, judgement-differentiation, complete wheat quality inspection robot
The orchestration of all steps.
The steering engine moves when putty at front and back chamber partition board and dredges putty.Feeding is controlled by aptitude manner, passes through biography
The information of sensor sensing, system can decide whether putty, whether empty material, Flow of Goods and Materials whether smooth homogeneous and can be according to these
Information carries out respective handling to ensure that feeding is continuous, stablize, uniformly.
With reference to Fig. 6, the transport mechanism 1-5 includes transparent synchrome conveying belt 1-5-1 and synchronous driving wheel 1-5-2, described
Synchronous driving wheel 1-5-2 is in rectangular layout on babinet 1-1, and the synchronous driving wheel 1-5-2 is equipped with synchronous gear I, described
The edge of bright synchrome conveying belt 1-5-1 is equipped with II 1-5-3 of synchronous gear, the synchronous gear II of the transparent synchrome conveying belt 1-5-1
I assembly connection of synchronous gear of 1-5-3 and synchronous driving wheel 1-5-2 partially pass through on the transparent synchrome conveying belt 1-5-1
Under two-sided machine visual mechanisms 1-6.
With reference to Fig. 7, the two-sided machine visual mechanisms 1-6 up and down includes upper machine vision mechanism 1-6-1, upper shadow-free light source
1-6-2, upper vision bottom plate 1-6-3, lower machine vision mechanism 1-6-4, lower shadow-free light source 1-6-5 and lower vision bottom plate 1-6-6, institute
Shu Shang machine vision mechanism 1-6-1, upper shadow-free light source 1-6-2 are located at rear on the upside of transparent synchrome conveying belt 1-5-1, it is described on regard
Feel that bottom plate 1-6-3 is set on the downside of transparent synchrome conveying belt 1-5-1, and immediately below upper shadow-free light source 1-6-2, the lower machine
Visual mechanisms 1-6-4, lower shadow-free light source 1-6-5 are located at front, the lower vision bottom plate 1-6-6 on the downside of transparent conveyor belt 1-5-1
On the upside of transparent synchrome conveying belt 1-5-1, and right over lower shadow-free light source 1-6-5.
Include sorting sensor 1-7-1, pulp-collecting box 1-7-2, air tube 1-7-3 and title with reference to Fig. 8, the mechanism for sorting 1-7
Sensor 1-7-4 is retransmitted, the sorting sensor 1-7-1 is mounted on the tops pulp-collecting box 1-7-2, and is located at transparent synchrome conveying belt 1-
Above 5-1, the air tube 1-7-3 is connected by the pipeline equipped with air valve with air pump, and on the transparent synchrome conveying belt 1-5-1 of face
Sample 1-7-5 to be picked, the weighing sensor 1-7-4 be mounted on the lower parts pulp-collecting box 1-7-2, realize sample 1-7-5's to be picked
It weighs.When work, sorting sensor 1-7-1 incudes sample 1-7-5 to be picked, and controlling air tube 1-7-3 by controller opens, and will wait for
It picks sample 1-7-5 and blows to pulp-collecting box 1-7-2, and weighed by weighing sensor 1-7-4.
Guide item is additionally provided in the present embodiment(It is not shown in figure), the guide item is located at feeding mechanism 1-2 and conveyer
Between structure 1-4.
Reference Fig. 9, a method of using the artificial intelligence wheat quality inspection robot to wheat progress quality inspection, including with
Lower step:
Wheat measuring samples are reached into feeding mechanism by the feeding mouth of artificial intelligence wheat quality inspection robot body, are passed through simultaneously
Sensor is responsible for detecting measuring samples in feeding mechanism and walks material state, and if there is putty situation, sensor is by putty information
Via controller sends system server to, controls voltage modulator by controller by system server, and then drive feeder
The vibrator of structure vibrates, and measuring samples are sent to transport mechanism for more than by feeding mechanism, wheat is waited for sample by transport mechanism
Product are sent to two-sided machine visual mechanisms up and down, are obtained by two-sided machine visual mechanisms up and down and are tested wheat samples original image,
It is former to every tested wheat samples by the artificial neural network system of the support graphics process built on system server again
Beginning image carries out algorithm process, and the conversion of the features such as the color of every tested wheat samples, texture, profile, size is characterized letter
Breath value inputs artificial neural network system, is compared with the character pair value of information stored on neuron, obtains character pair
The degree of membership of the value of information, then differentiated according to the degree of membership of the character pair value of information with S-shaped discriminant function, statistic of classification, lead to
Mechanism for sorting is crossed by unsound grain and impurity sorting shoulder to shoulder;Finally, the sample that detection is completed is output to discharging by transport mechanism
Mouthful, entire quality check process is all automatically performed.
The character pair value of information stored on the neuron obtains in the following manner:By sorted in advance 20000
Wheat samples(Including worm, bud, disease, damage, mould)It is sent to upper and lower Double-plate machine device visual mechanisms through feeding mechanism, transport mechanism, on
Under two-sided machine visual mechanisms obtain every wheat samples original image, pass through support graphics process artificial neural network system
Algorithm process is carried out to every wheat samples original image, converts the features such as the color of wheat, texture, profile, size to spy
The value of information is levied, and is stored on the neuron of artificial neural network system, a wheat characteristic information library is consequently formed.
Research practice shows using the present invention from wheat measuring samples information is obtained to detection judgement is completed, to each
The detection time of wheat samples is no more than 20 milliseconds, and a 50 grams of standard sample of artificial intelligence detection robot detection only needs 3-
Identification and data statistic analysis can be completed within 4 minutes.Whole process does not need human intervention, can effectively solve manually sense organ inspection
Survey the problem of by human factor and external environment influenced that Testing index is caused to drift about, meanwhile, robot to each sample all
It detects one by one, does not stay detection blind spot and dead angle.
The actually detected accuracy and repeatability index of the present invention:
Wheat unsound grain detects(By taking 50 grams of every part of sample of national standard as an example)
Whole unsound grains, error≤0.5%, repetitive error≤0.3%;
Injured kernel, error≤0.5%, repetitive error≤0.3%;
Germinate grain, error≤0.5%, repetitive error≤0.3%;
Scab grain(Red mould grain), error≤0.3%, repetitive error≤0.2%;
Scab grain(Black embryo), error≤0.2%, repetitive error≤0.1%;
Broken kernel(It is broken), error≤0.2%, repetitive error≤0.1%;
Moldy kernel, error≤0.3%, repetitive error≤0.2%;
Yellow rice kernel, every 1000 errors are less than 1, repetitive error zero.
Claims (9)
1. artificial intelligence wheat quality inspection robot, including artificial intelligence wheat quality inspection robot body and system server;It is described
Artificial intelligence wheat quality inspection robot body includes babinet, and the babinet is equipped with feeding mouth and discharge port, is set in the babinet
There are feeding mechanism, transport mechanism, upper and lower two-sided machine visual mechanisms and mechanism for sorting, the feeding mouth to be connected with feeding mechanism
Logical, the feeding mechanism is connected with transport mechanism, and the transport mechanism passes through upper and lower two-sided machine visual mechanisms, the transmission
Mechanism is connected with discharge port, and the mechanism for sorting is located at upper and lower two-sided machine visual mechanisms rear, the two-sided machine up and down
Visual mechanisms, mechanism for sorting are connected with system server;It is characterized in that:The feeding mechanism includes feeding box, the feeding
It is equipped with a partition board in box, the cavity of feeding box is divided into ante-chamber and back cavity, the feeding box bottom is equipped with one and runs through ante-chamber with after
The hopper of chamber, the back cavity are equipped with whether monitoring back cavity has the sensor I of material, the back cavity to be equipped with one close to baffle material trough rim
Horizontal direction monitors that the sensor II of material position, hopper center of the ante-chamber at partition board are equipped with what whether monitoring back cavity discharged
Sensor III, ante-chamber hopper exit upper edge be equipped with monitoring ante-chamber whether the sensor IV of putty, set on the feeding box
Having prevents the mechanism of back cavity putty, and the feeding box lower part is equipped with vibrator, and the vibrator is connected with voltage modulator, described
Sensor I, sensor III, sensor IV, prevents the mechanism of back cavity putty, voltage modulator to be connected with controller at sensor II,
The controller is connected with system server.
2. artificial intelligence wheat quality inspection robot according to claim 1, it is characterised in that:The horizontal direction monitoring material
Position sensor II be located at back cavity horizontal direction distance bottom >=1cm at.
3. artificial intelligence wheat quality inspection robot according to claim 1 or 2, it is characterised in that:It is described to prevent back cavity stifled
The mechanism of material includes steering engine, and the steering engine is located at outside feeding box, and the steering engine is connected with steering engine plectrum, the rudder by steering engine connecting rod
Machine plectrum is located in back cavity, and close to partition board and hopper.
4. artificial intelligence wheat quality inspection robot according to claim 1 or 2, it is characterised in that:The transport mechanism packet
Transparent synchrome conveying belt and synchronous driving wheel are included, the synchronous driving wheel is arranged on babinet, and the synchronous driving wheel is equipped with
Synchronous gear I, the edge of the transparent synchrome conveying belt are equipped with synchronous gear II, the synchronous gear II of the transparent synchrome conveying belt with
I assembly connection of synchronous gear of synchronous driving wheel partially passes through two-sided machine vision machine up and down on the transparent synchrome conveying belt
Structure.
5. artificial intelligence wheat quality inspection robot according to claim 4, it is characterised in that:The two-sided machine up and down regards
Feel mechanism include upper machine vision mechanism, upper shadow-free light source, upper vision bottom plate, lower machine vision mechanism, lower shadow-free light source and under
Vision bottom plate, the upper machine vision mechanism, upper shadow-free light source are located at rear on the upside of transparent synchrome conveying belt, the upper vision bottom
Plate is set on the downside of transparent synchrome conveying belt, and immediately below upper shadow-free light source, the lower machine vision mechanism, lower shadow-free light source
The front on the downside of transparent conveyor belt, the lower vision bottom plate are set on the upside of transparent synchrome conveying belt, and are located at lower shadow-free light source
Surface.
6. artificial intelligence wheat quality inspection robot according to claim 5, it is characterised in that:The mechanism for sorting includes point
Sensor, pulp-collecting box, air tube and weighing sensor are picked, the sorting sensor is mounted on pulp-collecting box top, and positioned at transparent same
It walks above conveyer belt, the air tube is connected by the pipeline equipped with air valve with air pump, and waiting for of taking of the transparent synchronous driving of face
Sample is picked, the weighing sensor is mounted on pulp-collecting box lower part, realizes weighing for sample to be picked.
7. artificial intelligence wheat quality inspection robot according to claim 1 or 2, it is characterised in that:It is additionally provided with guide item, institute
Guide item is stated between feeding mechanism and transport mechanism.
8. a kind of method carrying out quality inspection to wheat using artificial intelligence wheat quality inspection robot described in claim 1, feature
It is, includes the following steps:
Wheat measuring samples are reached into feeding mechanism by the feeding mouth of artificial intelligence wheat quality inspection robot body, are passed through simultaneously
Measuring samples walk material state in sensor detection feeding mechanism, and if there is putty situation, sensor is by putty information through control
Device processed sends system server to, controls voltage modulator by controller by system server, and then drive feeding mechanism
Vibrator vibrates, and measuring samples are sent to transport mechanism for more than by feeding mechanism, are sent wheat measuring samples by transport mechanism
To upper and lower two-sided machine visual mechanisms, is obtained by two-sided machine visual mechanisms up and down and be tested wheat samples original image, then led to
The artificial neural network system of the support graphics process built on system server is crossed to every tested wheat samples original graph
As carrying out algorithm process, converts the feature of every tested wheat samples to characteristic information value, inputs artificial neural network system,
It is compared with the character pair value of information stored on neuron, obtains the degree of membership of the character pair value of information, then with differentiating letter
Several degrees of membership according to the character pair value of information are differentiated that statistic of classification is by unsound grain and miscellaneous shoulder to shoulder by mechanism for sorting
Matter sorts;Finally, the sample that detection is completed is output to discharge port by transport mechanism, entire quality check process is all automatically performed.
9. the method that the artificial intelligence wheat quality inspection robot according to claim 8 carries out quality inspection to wheat, special
Sign is that the character pair value of information stored on the neuron obtains in the following manner:
Wheat samples sorted in advance are sent through feeding mechanism, transport mechanism to upper and lower Double-plate machine device visual mechanisms, up and down
Two-sided machine visual mechanisms obtain every wheat samples original image, by the artificial neural network system couple for supporting graphics process
Every wheat samples original image carries out algorithm process, converts the feature of wheat to characteristic information value, and be stored in artificial god
On neuron through network system, a wheat characteristic information library is consequently formed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810830624.4A CN108465644B (en) | 2018-07-26 | 2018-07-26 | Artificial intelligent wheat quality inspection robot and quality inspection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810830624.4A CN108465644B (en) | 2018-07-26 | 2018-07-26 | Artificial intelligent wheat quality inspection robot and quality inspection method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108465644A true CN108465644A (en) | 2018-08-31 |
CN108465644B CN108465644B (en) | 2020-06-02 |
Family
ID=63259898
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810830624.4A Active CN108465644B (en) | 2018-07-26 | 2018-07-26 | Artificial intelligent wheat quality inspection robot and quality inspection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108465644B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111359907A (en) * | 2020-02-24 | 2020-07-03 | 南京农业大学 | Real-time online detection system and method for wheat scab infection particle rate based on embedded deep learning |
CN111389747A (en) * | 2020-04-26 | 2020-07-10 | 成都复星联创科技有限公司 | Bamboo strip intelligent sorting system and method based on visual identification |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
SU971524A1 (en) * | 1981-04-09 | 1982-11-07 | Центральный научно-исследовательский институт оловянной промышленности | Self-oscillating separator |
CN202070498U (en) * | 2011-04-15 | 2011-12-14 | 伟钧精密机械有限公司 | Object detector |
JP2012229118A (en) * | 2011-04-27 | 2012-11-22 | Sinfonia Technology Co Ltd | Article sorting device |
CN104475362A (en) * | 2014-12-08 | 2015-04-01 | 浙江大学 | Real-time red date detecting and grading device based on machine vision technology |
CN106743191A (en) * | 2017-03-27 | 2017-05-31 | 重庆市永川区新窝建材有限公司 | A kind of oscillation feed gear |
CN207507850U (en) * | 2017-02-08 | 2018-06-19 | 云南万视智能设备有限公司 | A kind of crawler type color sorting machine |
-
2018
- 2018-07-26 CN CN201810830624.4A patent/CN108465644B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
SU971524A1 (en) * | 1981-04-09 | 1982-11-07 | Центральный научно-исследовательский институт оловянной промышленности | Self-oscillating separator |
CN202070498U (en) * | 2011-04-15 | 2011-12-14 | 伟钧精密机械有限公司 | Object detector |
JP2012229118A (en) * | 2011-04-27 | 2012-11-22 | Sinfonia Technology Co Ltd | Article sorting device |
CN104475362A (en) * | 2014-12-08 | 2015-04-01 | 浙江大学 | Real-time red date detecting and grading device based on machine vision technology |
CN207507850U (en) * | 2017-02-08 | 2018-06-19 | 云南万视智能设备有限公司 | A kind of crawler type color sorting machine |
CN106743191A (en) * | 2017-03-27 | 2017-05-31 | 重庆市永川区新窝建材有限公司 | A kind of oscillation feed gear |
Non-Patent Citations (1)
Title |
---|
王耀南: "《计算机图像处理与识别技术》", 30 June 2001, 北京:高等教育出版社 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111359907A (en) * | 2020-02-24 | 2020-07-03 | 南京农业大学 | Real-time online detection system and method for wheat scab infection particle rate based on embedded deep learning |
CN111359907B (en) * | 2020-02-24 | 2021-08-20 | 南京农业大学 | Real-time online detection system and method for wheat scab infection particle rate based on embedded deep learning |
CN111389747A (en) * | 2020-04-26 | 2020-07-10 | 成都复星联创科技有限公司 | Bamboo strip intelligent sorting system and method based on visual identification |
Also Published As
Publication number | Publication date |
---|---|
CN108465644B (en) | 2020-06-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108469435A (en) | Artificial intelligence rice quality inspection robot and quality detecting method | |
CN108465649A (en) | Artificial intelligence corn quality inspection robot and quality detecting method | |
EP3408640B1 (en) | Methods and apparatus adapted to identify a specimen container from multiple lateral views | |
US4884696A (en) | Method and apparatus for automatically inspecting and classifying different objects | |
US11815446B2 (en) | Methods and apparatus for characterizing a specimen container and specimen | |
US9024223B2 (en) | Optical type granule sorting machine | |
CN102217559B (en) | On-line automatic detection and sorting equipment and method for incubated egg quality | |
US8154593B2 (en) | Appearance inspection device | |
CN106290382A (en) | Bubble-cap tablet package defective vision detection device and method | |
US20100214560A1 (en) | Carrying device and appearance inspection device for test objects | |
CN104634790A (en) | Capsule detection method and high-speed fully-automatic detection device | |
US5903341A (en) | Produce grading and sorting system and method | |
CN106546173A (en) | For detecting the equipment and its detection method of components and parts | |
CN108465644A (en) | Artificial intelligence wheat quality inspection robot and quality detecting method | |
WO2018089935A1 (en) | Methods and apparatus for characterizing a specimen using pattern illumination | |
CN109433641A (en) | The filling omission intelligent detecting method of tablet capsule based on machine vision | |
CN102735694A (en) | Automatic capsule detection machine | |
CN103406279A (en) | Intelligent three-camera visual fruit detection device | |
CN202133655U (en) | Online automatic detecting sorting equipment for quality of incubated eggs | |
CN112077026A (en) | Imperfect grain detection equipment | |
CN108204977A (en) | Corn kernel quality automatic detection device based on machine vision | |
JP2000180369A (en) | Method and apparatus for measurement of appearance quality of grain | |
JP2002346483A (en) | Grain component analyzer | |
JP5274706B1 (en) | Sorting set system and sorting set method | |
CN113128502A (en) | Single piece separation system, method and device based on deep learning and machine vision |
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 |