CN101673344A - Device and method for recognizing mature oranges in natural scene by filter plate spectral image technology - Google Patents
Device and method for recognizing mature oranges in natural scene by filter plate spectral image technology Download PDFInfo
- Publication number
- CN101673344A CN101673344A CN200910232914A CN200910232914A CN101673344A CN 101673344 A CN101673344 A CN 101673344A CN 200910232914 A CN200910232914 A CN 200910232914A CN 200910232914 A CN200910232914 A CN 200910232914A CN 101673344 A CN101673344 A CN 101673344A
- Authority
- CN
- China
- Prior art keywords
- spectrum
- filter plate
- oranges
- tangerines
- under
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000005516 engineering process Methods 0.000 title claims abstract description 24
- 230000003595 spectral effect Effects 0.000 title claims abstract description 15
- 235000013399 edible fruits Nutrition 0.000 claims abstract description 29
- 238000001228 spectrum Methods 0.000 claims description 72
- 241000675108 Citrus tangerina Species 0.000 claims description 49
- 238000001914 filtration Methods 0.000 claims description 11
- 239000000284 extract Substances 0.000 claims description 10
- 206010061274 Malocclusion Diseases 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 7
- 238000005192 partition Methods 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 claims description 2
- 238000003306 harvesting Methods 0.000 abstract description 4
- 238000005286 illumination Methods 0.000 description 5
- 239000000463 material Substances 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 241000219307 Atriplex rosea Species 0.000 description 2
- 239000004744 fabric Substances 0.000 description 2
- 229920006395 saturated elastomer Polymers 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 239000004575 stone Substances 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/46—Measurement of colour; Colour measuring devices, e.g. colorimeters
- G01J3/50—Measurement of colour; Colour measuring devices, e.g. colorimeters using electric radiation detectors
- G01J3/51—Measurement of colour; Colour measuring devices, e.g. colorimeters using electric radiation detectors using colour filters
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
- G06V10/12—Details of acquisition arrangements; Constructional details thereof
- G06V10/14—Optical characteristics of the device performing the acquisition or on the illumination arrangements
- G06V10/143—Sensing or illuminating at different wavelengths
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/68—Food, e.g. fruit or vegetables
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Image Analysis (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a device and a method for recognizing mature oranges in a natural scene based on a filter plate spectral image technology and relates to a device and a method for recognizing oranges and branches in a natural scene. The device consists of a CCD camera (1), a pan tilt (2), a filter plate (5), a camera tripod (8), a baffle (3), a turntable (4) and a computer (10); wherein, the CCD camera (1) can sense near infrared signal, the pan tilt (2) is mounted on the camera tripod (8), the baffle (3) has round hole and is fixed vertically on the pan tilt (2) of the camera tripod (8), the CCD camera (1) is arranged on one side of the baffle (3) and a camera lens is embedded in the round hole of the baffle (3), the turntable (4) is fixed on the other side of the baffle (3), the turntable (4) has a round hole and can rotate freely, and the filter plate (5) is arranged in the hole of the turntable (4) respectively. The device and the method lay a foundation for developing intelligent fruit harvesting robots and popularization thereof.
Description
Technical field
The present invention relates to the recognition device and the method thereof of oranges and tangerines and branch under a kind of natural scene, refer in particular to devices and methods therefor based on oranges and tangerines and branch under the filter plate spectrum picture technology identification natural scene.
Technical background
Present traditional fruit picking methods such as hand is plucked, cutter is cut, pole is disclosed cause fruit damage in various degree easily, cause occurring rotting at storage and selling period.Traditional in addition fruit picking is that intensity is big, the work of length consuming time.So exploitation possesses, and the intelligent robot of reliable operation replaces artificial the harvesting to have crucial meaning under the natural scene.The operate as normal of fruit picking robot all depends on the correct identification to fruit.Thereby to realize the results of fruit picking robot to fruit, key is to identify fruit under the natural scene of complexity.But fruit tree comes in every shape, uneven, many fruits are positioned between branch back or two branches, and the branch of fruit tree is thicker, if colliding, mechanical arm and these branches will certainly damage mechanical arm, desire to make mechanical arm can pluck these fruits, therefore must avoid branch, equally also will identify barrier such as branch so that, finish the harvesting of fruit for the motion of mechanical arm provides parameter.
By retrieval, do not see the relevant patent of fruit identification under the natural scene, only retrieve the related invention of the robot industry application object identification under structural environment, as " device and method with industrial robot of coloured image recognition capability; patent No. 200410093502.X ", " image recognition apparatus, method and robot device, application number 200710138345.3 " etc.Foregoing invention mainly is to satisfy the job requirements of robot under structural industrial environment, environment is simple, stable, not disturbed by factor such as illumination, therefore only need mate, can reach the purpose of detected object by the model image in object images and the database.And under the physical environment of unstructuredness, because the difference of fruit size and growing space, can not there be a definite model image and object images to realize coupling, and the changeability of direction of illumination and intensity, the influence of factors such as the complicacy of background certainly will also can cause difficulty to identification.At the complicated physical environment of unstructuredness, must seek a kind of industrial circle of distinguishing, the new method of accuracy of identification height, strong robustness.The inside organic principle difference of other background material such as fruit, leaf, branch and weeds, their spectral characteristic is also inequality, and the absorption value to light under different-waveband also there are differences.Therefore the spectrum picture technology can integrate image information and spectral information, and spectral information that can the analytic target image from bulk properties such as constituents, reaches the purpose of effective identification fruit and branch.
Summary of the invention
The objective of the invention is at fruit picking robot, designed a kind of based on the fruit of filter plate spectrum picture technology and the recognition methods of branch thereof, for the motion of mechanical arm provides parameter.
The objective of the invention is to realize by the following method: the device of discerning ripe oranges and tangerines under the natural scene based on filter plate spectrum picture technology: mainly form by CCD camera, The Cloud Terrace, filter plate, camera trivets, baffle plate, rotating disk and computing machine; It is characterized in that the CCD camera can respond near infrared signal, The Cloud Terrace is installed on the camera trivets, and baffle plate has circular hole and is vertically fixed on the The Cloud Terrace of camera trivets; In a side of baffle plate, lay the CCD camera, and its camera lens is embedded the circular hole of baffle plate; At the opposite side of baffle plate, rotating disk is fixed on the baffle plate, rotating disk has five circular holes and can freely rotate, and settles filter plate respectively in the circular hole of this rotating disk.
Method based on ripe oranges and tangerines under the filter plate spectrum picture technology identification natural scene: at first, collect the filtering image under the different-waveband Same Scene by above-mentioned device based on ripe oranges and tangerines under the filter plate spectrum picture technology identification natural scene; And with data transmission in computing machine, make up the 3 D stereo data block, then this data block is carried out pre-service; To the characteristic of pretreated data extract fruit and branch, can reach the purpose of ripe oranges and tangerines under the identification natural scene at last.
The collection of described view data: at first adjust the parameter (time shutter, aperture size etc.) of camera, guarantee to collect distinct image; By the rotation of rotating disk, when the scale below the circular hole on the scale in the filter plate outside on the rotating disk and the baffle plate aligns when overlapping, promptly represent the camera lens of camera, the circular hole of baffle plate and the complete being aligned of filter plate three on the rotating disk just can carry out image acquisition.So both can guarantee the stable of camera, also can guarantee between camera lens and the filter plate tightly not light leak.By the rotation of rotating disk, just can collect the filtering image under the different wave length under the Same Scene.
The pre-service of described raw data: can make up the three-dimensional data piece by the filtering image that collects, in this data block, each pixel all has a gray-scale value under different band images, connect these gray-scale values, just formed a curve of spectrum, promptly each pixel is all corresponding to a curve of spectrum.Because light signal filters via filter plate, the influence of the factors such as existence of CCD camera dark noise, will certainly there be suitable noise in the three-dimensional data piece that is made up by filtering image, therefore is necessary to carry out pre-service, to remove noise.The minimal noise partition method can be removed noise signal effectively, extracts valuable information, and can enlarge the SPECTRAL DIVERSITY between oranges and tangerines, branch, leaf and background material, therefore at first raw data is carried out the minimal noise separating treatment.
The extraction of described characteristic: the classification of application of spectral angle is carried out feature extraction to pretreated data; Spectrum angle classification is a kind of matching technique of spectrum, and this technology is based on estimating the spectrum to be measured and the curve of spectrum of distinguishing each pixel point with reference to spectral similarity; Said method need provide oranges and tangerines and branch with reference to spectrum; In any surface finish, choose area-of-interest (ROI) on the sample image of representative oranges and tangerines and branch, and calculate the spectrum mean value that comprises pixel among the ROI, promptly get oranges and tangerines and branch with reference to spectrum.Spectrum angle classification is spectrum to be measured by each pixel relatively and extract the feature of oranges and tangerines and branch with reference to the difference between spectrum just.
The advantage that the present invention compares with conventional art:
(1) it is complete to obtain information.The spectrum picture technology both can have been obtained the band image information of object, also can obtain the spectral information of object simultaneously.
(2) analytical approach is unique novel.Traditional recognition methods mainly is the technology that adopts based on local gray level, color characteristic and shape facility, and generally is confined to carry out graphical analysis on visible light and the two-dimensional image information category, excavates less than more information.The spectrum picture technology then can be from three-dimensional information, utilize classification this unique novel method in spectrum angle to analyze spectral information, excavates the information that multipotency is more discerned oranges and tangerines and branch enough effectively.
(3) accuracy of identification height.The spectrum picture technology is the difference according to the spectral characteristic of fruit, branch, leaf and other background material, analyzes the feature of extracting object from the bulk properties of object, can effectively overcome influences such as uneven illumination is even, complex background, the accuracy of identification height.
(4) strong robustness.Operation is plucked by robot under the physical environment of complexity, so the factors such as uncertainty of the changeability of intensity of illumination and direction, complex background all can cause the weak robustness of traditional recognition method.And the spectrum picture technology is to be starting point with the curve of spectrum, and is uncorrelated with the intensity of image, can react the bulk properties of object simultaneously, therefore can overcome the influence of illumination and complex background, strong robustness.
The invention has the beneficial effects as follows:
Utilize optical filter spectrum picture technology can be for the fruit tree image under the Analysis of Complex natural scene time, a kind of novel identification technology is provided, opened up a kind of analysis means of frontier, broken the only limitation of analysis image on two-dimentional aspect of classic method.This technology is in conjunction with the bulk properties and the surface of manipulating object, and utilize spectrum angle classification, from the three-dimensional data piece that collects, excavate how valuable information, improved accuracy of identification, and has good robustness at different physical environments, the reliable and stable intelligent fruit harvest machine people of operation lays the first stone under recognition capability, visual performance, light and soft softization of operation, the mal-condition for exploitation has, and also provides possibility for the practicability of promoting fruit picking robot.
Description of drawings
Fig. 1: invention application example implement device synoptic diagram.
Wherein, 1, CCD camera; 2, The Cloud Terrace; 3, baffle plate; 4, rotating disk; 5, filter plate; 6, the scale mark on the baffle plate; 7, the scale mark on the rotating disk; 8, camera trivets; 9, fruit tree; 10, computing machine.
Fig. 2: the process flow diagram of the inventive method
Embodiment
It is example that the present invention extracts characteristics of objects with spectrum angle sorting algorithm, introduces the concrete embodiment of the present invention.
Invention application example realization hardware synoptic diagram is consulted Fig. 1, and process flow diagram of the present invention is consulted Fig. 2.The CCD camera of the camera that example is selected for use near infrared region is had good inductive effects.At first baffle plate 3 is vertically fixed on the tripod The Cloud Terrace 2, and CCD camera 1 is placed in a side of baffle plate 3, the circular hole with its alignment lens baffle plate 3 prevents light leak; The circular hole that then five filter plates 5 is embedded into rotating disk 4 respectively, and be fixed on the opposite side of baffle plate 3 at this rotating disk, guaranteeing to avoid light leak, rotating disk 4 can freely rotate simultaneously; When the red scale mark 7 in filter plate 5 outsides on the rotating disk 4 and red scale mark 6 on the baffle plate 3 align when overlapping, promptly show the camera lens of camera 1, circular hole and filter plate 5 three's being aligneds of baffle plate 3, this moment can images acquired.
At first adjust the time shutter of camera, parameters such as aperture size make camera can collect distinct image, by rotating rotating disk 4, gather the filtering image of five width of cloth oranges and tangerines fruit trees of different-waveband under the Same Scene; After the data that collect are input to computing machine, make up the three-dimensional data piece; Then use the minimal noise partition method that raw data is carried out pre-service, with the removal noise, and the SPECTRAL DIVERSITY between expansion oranges and tangerines, branch, leaf and background material; On saturated oranges and tangerines of any surface finish, color and branch, choose area-of-interest (ROI) respectively then, and calculate the averaged spectrum of all pixels in this zone, constitute oranges and tangerines and branch with reference to spectrum; Utilize spectrum angle sorting algorithm pretreated data to be carried out the feature extraction of oranges and tangerines and branch at last.
As shown in Figure 2, the concrete grammar of extraction oranges and tangerines and branch feature is divided into two stages: " modelling " (frame of broken lines on Fig. 2 left side) and " Model Matching " (frame of broken lines on Fig. 2 the right)." modelling " is the process that training image is trained, and promptly is identified for the oranges and tangerines of spectrum angle classification and the process with reference to spectrum and spectrum angle threshold value of branch." Model Matching " is the process of testing image being mated identification, promptly extracts the process of the feature of oranges and tangerines and branch.
The concrete steps of " modelling " are as follows:
(1) gather filtering image: at first adjust the time shutter of camera, parameters such as aperture size make camera can collect distinct image, by rotating rotating disk 4, gather the filtering image of five width of cloth oranges and tangerines fruit trees of different-waveband under the Same Scene.
(2) make up the three-dimensional data piece: after the data that collect are input to computing machine, make up the three-dimensional data piece.In the three-dimensional data piece, contain the image information of different-waveband, also comprise the spectral information of each pixel simultaneously.
(3) data pre-service: utilization minimal noise partition method is carried out pre-service to raw data, with the removal noise, and the SPECTRAL DIVERSITY between expansion oranges and tangerines, branch, leaf and background material.
(4) choose the ROI of oranges and tangerines and branch: whether area-of-interest (ROI) oranges and tangerines that comprised or the character of branch is single, be whether selected ROI only contains a kind of in oranges and tangerines and the branch, the representativeness that whether has oranges and tangerines or branch, this directly has influence on the result of last feature extraction.Therefore, on saturated oranges and tangerines of any surface finish, color or branch, choose ROI.
(5) calculate the oranges and tangerines branch with reference to spectrum: calculate the averaged spectrum of all pixels in the ROI that (4) set by step require to choose, constitute oranges and tangerines and branch with reference to spectrum.
(6) extract the spectrum to be measured of each pixel: the spectrum to be measured that extracts each pixel in the three-dimensional data piece.So-called spectrum to be measured is the pixel curve of spectrum that pairing brightness value (gray-scale value) is formed under each wave band.
(7) carry out spectrum angle classification: utilization spectrum angle sorting algorithm calculate oranges and tangerines and branch with reference to the spectrum generalized angle between spectrum and each the pixel curve of spectrum, obtain the angular image of oranges and tangerines and branch.
(8) determine the spectrum angle threshold value of oranges and tangerines and branch: the angular image that is obtained by step (7) is made repeated attempts to be shown, when spectrum angle radian threshold value is in [0.15-0.24] is interval, the recognition effect ideal of oranges and tangerines, when spectrum angle radian threshold value gets 0.2, therefore recognition effect the best determines that it is the spectrum angle threshold value of oranges and tangerines; When spectrum angle radian threshold value is in [0.07-0.18] is interval, the recognition effect ideal of branch, when spectrum angle radian threshold value gets 0.12, branch recognition effect the best, so determine that it is the spectrum angle threshold value of oranges and tangerines.
Therefore, 8 steps by above-mentioned " modelling " obtained being used for the oranges and tangerines of " Model Matching " and branch with reference to spectrum and spectrum angle threshold value.
The same with " modelling ", " Model Matching " stage at first also will gather filtering image, make up the three-dimensional data piece, carry out the data pre-service and extract the spectrum to be measured of each pixel, and its method all method with " modelling " is identical; Utilize then spectrum angle sorting algorithm in conjunction with oranges and tangerines of determining in " modelling " and branch with reference to spectrum and spectrum angle threshold value, can extract the feature of the oranges and tangerines and the branch of object images under the different scenes.
Claims (6)
1, discerns the device of ripe oranges and tangerines under the natural scene based on filter plate spectrum picture technology: mainly form by CCD camera (1), The Cloud Terrace (2), filter plate (5), camera trivets (8), baffle plate (3), rotating disk (4) and computing machine (10); It is characterized in that CCD camera (1) can respond near infrared signal, The Cloud Terrace (2) is installed on the camera trivets (8), and baffle plate (3) has circular hole and is vertically fixed on the The Cloud Terrace (2) of camera trivets (8); In a side of baffle plate (3), lay CCD camera (1), and its camera lens is embedded the circular hole of baffle plate (3); Opposite side in baffle plate (3) is fixed on (3) on the baffle plate with rotating disk (4), and rotating disk (4) has circular hole and can freely rotate, and settles filter plate (5) in the circular hole of this rotating disk (4) respectively.
2, the device based on ripe oranges and tangerines under the filter plate spectrum picture technology identification natural scene according to claim 1, it is characterized in that filter plate (5) is visible light and near infrared filter plate, its characteristic wavelength is respectively: 440nm, 550nm, 658nm, 790nm and 850nm.
3, based on the method for ripe oranges and tangerines under the filter plate spectrum picture technology identification natural scene, it is characterized in that: computing machine (10) at first collects the filtering image under the different-waveband Same Scene; And with data transmission in the computing machine (10), make up the 3 D stereo data block, then this data block is carried out pre-service; Extract the characteristic of fruit and branch at last from pretreated data block, can discern ripe oranges and tangerines under the natural scene.
4, the method based on ripe oranges and tangerines under the filter plate spectrum picture technology identification natural scene according to claim 3, it is characterized in that the collection of described view data: at first adjust time shutter, the aperture size parameter of CCD camera (1), guarantee to collect distinct image; Rotation by rotating disk (4), going up the scale in filter plate (5) outside and scale below baffle plate (3) is gone up circular hole when rotating disk (4) aligns when overlapping, the camera lens of promptly representing CCD camera (1), the complete being aligned of filter plate (5) three on the circular hole of baffle plate (3) and the rotating disk (4) just can carry out image acquisition; By the rotation of rotating disk (4), just can collect the filtering image under the different wave length under the Same Scene.
5, the method based on ripe oranges and tangerines under the filter plate spectrum picture technology identification natural scene according to claim 3, it is characterized in that the pre-service of described raw data: make up the three-dimensional data piece by the filtering image that collects, in this data block, each pixel all has a gray-scale value under different band images, connect these gray-scale values, just formed a curve of spectrum, promptly each pixel is all corresponding to a curve of spectrum; Use the minimal noise partition method raw data is carried out the minimal noise separating treatment.
6, the method based on ripe oranges and tangerines under the filter plate spectrum picture technology identification natural scene according to claim 3, it is characterized in that the extraction of described characteristic: application of spectral angle classification is carried out feature extraction to pretreated data; Earlier in any surface finish, choosing area-of-interest on the sample image of representative oranges and tangerines and branch is ROI, and calculates the spectrum mean value that comprises pixel among the ROI, promptly get oranges and tangerines and branch with reference to spectrum; Adopt spectrum angle classification by spectrum to be measured that compares each pixel and the characteristic of extracting oranges and tangerines and branch with reference to the difference between spectrum again.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910232914A CN101673344A (en) | 2009-10-09 | 2009-10-09 | Device and method for recognizing mature oranges in natural scene by filter plate spectral image technology |
| PCT/CN2009/001537 WO2011041924A1 (en) | 2009-10-09 | 2009-12-23 | Device and method for identifying ripe oranges in nature scene by filter spectral image technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910232914A CN101673344A (en) | 2009-10-09 | 2009-10-09 | Device and method for recognizing mature oranges in natural scene by filter plate spectral image technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101673344A true CN101673344A (en) | 2010-03-17 |
Family
ID=42020565
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910232914A Pending CN101673344A (en) | 2009-10-09 | 2009-10-09 | Device and method for recognizing mature oranges in natural scene by filter plate spectral image technology |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN101673344A (en) |
| WO (1) | WO2011041924A1 (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104794442A (en) * | 2015-04-16 | 2015-07-22 | 王爱云 | Platform for identifying ripeness degrees of integral apple trees |
| CN106124049A (en) * | 2016-06-20 | 2016-11-16 | 福州大学 | A kind of implementation method of Vegetation canopy multi-optical spectrum imaging system |
| CN106407962A (en) * | 2016-11-15 | 2017-02-15 | 融安县植保植检站 | Citrus fruit fly harm fruit-drop rate statistics system |
| CN107014814A (en) * | 2017-05-25 | 2017-08-04 | 河南嘉禾智慧农业科技有限公司 | A kind of fruit maturity automatic recognition system |
| CN108773332A (en) * | 2018-04-28 | 2018-11-09 | 贵州维迪话科技有限公司 | A kind of vehicle monitoring system |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9462749B1 (en) | 2015-04-24 | 2016-10-11 | Harvest Moon Automation Inc. | Selectively harvesting fruits |
| US9468152B1 (en) | 2015-06-09 | 2016-10-18 | Harvest Moon Automation Inc. | Plant pruning and husbandry |
| US9965845B2 (en) | 2016-07-11 | 2018-05-08 | Harvest Moon Automation Inc. | Methods and systems for inspecting plants for contamination |
| US9928584B2 (en) | 2016-07-11 | 2018-03-27 | Harvest Moon Automation Inc. | Inspecting plants for contamination |
| CN106199358A (en) * | 2016-08-15 | 2016-12-07 | 哈尔滨理工大学 | A kind of electric branch image acquisition and method of synthesis in polymer |
| CN111325767B (en) * | 2020-02-17 | 2023-06-02 | 杭州电子科技大学 | Real scene-based citrus fruit tree image set synthesis method |
| CN111798433B (en) * | 2020-07-08 | 2024-04-26 | 贵州师范大学 | Method for identifying and counting mature dragon fruits in mountain area of plateau based on unmanned aerial vehicle remote sensing |
| CN117347312B (en) * | 2023-12-06 | 2024-04-26 | 华东交通大学 | Orange continuous detection method and equipment based on multispectral structured light |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2606573B1 (en) * | 1986-11-06 | 1993-08-06 | Cemagref | METHOD AND DEVICE FOR REAL-TIME ANALOGUE SPECTRAL PRESELECTION, FOR EXAMPLE FOR ARTIFICIAL VISION SYSTEM |
| JPH05174131A (en) * | 1991-12-25 | 1993-07-13 | Iseki & Co Ltd | Visual device for fruit harvest robot |
| CN2636279Y (en) * | 2003-07-26 | 2004-08-25 | 鸿富锦精密工业（深圳）有限公司 | Viewfinder |
| CN101013091B (en) * | 2007-01-25 | 2011-06-01 | 江西农业大学 | Method and device for non-invasive detecting of soil and pesticide contamination on fruit surface |
-
2009
- 2009-10-09 CN CN200910232914A patent/CN101673344A/en active Pending
- 2009-12-23 WO PCT/CN2009/001537 patent/WO2011041924A1/en active Application Filing
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104794442A (en) * | 2015-04-16 | 2015-07-22 | 王爱云 | Platform for identifying ripeness degrees of integral apple trees |
| CN104794442B (en) * | 2015-04-16 | 2017-11-14 | 桂林电子科技大学 | Whole strain apple tree cooking level Identification platform |
| CN106124049A (en) * | 2016-06-20 | 2016-11-16 | 福州大学 | A kind of implementation method of Vegetation canopy multi-optical spectrum imaging system |
| CN106407962A (en) * | 2016-11-15 | 2017-02-15 | 融安县植保植检站 | Citrus fruit fly harm fruit-drop rate statistics system |
| CN107014814A (en) * | 2017-05-25 | 2017-08-04 | 河南嘉禾智慧农业科技有限公司 | A kind of fruit maturity automatic recognition system |
| CN108773332A (en) * | 2018-04-28 | 2018-11-09 | 贵州维迪话科技有限公司 | A kind of vehicle monitoring system |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2011041924A1 (en) | 2011-04-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101673344A (en) | Device and method for recognizing mature oranges in natural scene by filter plate spectral image technology | |
| Nuske et al. | Automated visual yield estimation in vineyards | |
| Song et al. | Kiwifruit detection in field images using Faster R-CNN with VGG16 | |
| Hannan et al. | A machine vision algorithm combining adaptive segmentation and shape analysis for orange fruit detection | |
| Zhou et al. | Using colour features of cv.‘Gala’apple fruits in an orchard in image processing to predict yield | |
| EP2548147B1 (en) | Method to recognize and classify a bare-root plant | |
| Kurtser et al. | In-field grape cluster size assessment for vine yield estimation using a mobile robot and a consumer level RGB-D camera | |
| Wang et al. | Deep learning approach for apple edge detection to remotely monitor apple growth in orchards | |
| CN110765916B (en) | Farmland seedling ridge identification method and system based on semantics and example segmentation | |
| Malik et al. | Detection and counting of on-tree citrus fruit for crop yield estimation | |
| CN104751199B (en) | A kind of cotton splits bell phase automatic testing method | |
| KR20150000435A (en) | Recongnition of Plant Growth Steps and Environmental Monitoring System and Method thereof | |
| CN104050668A (en) | Object recognition method applied to green tea tender shoots and based on binocular vision technology | |
| Ge et al. | Three dimensional apple tree organs classification and yield estimation algorithm based on multi-features fusion and support vector machine | |
| Arefi et al. | Development of an expert system based on wavelet transform and artificial neural networks for the ripe tomato harvesting robot | |
| Changyi et al. | Apple detection from apple tree image based on BP neural network and Hough transform | |
| Kuznetsova et al. | Detecting apples in orchards using YOLOv3 | |
| CN103279762A (en) | Judging method for common growth form of fruit under natural environment | |
| CN109115719A (en) | A kind of Citrus Huanglongbing pathogen Band fusion rapid detection method based on high light spectrum image-forming technology | |
| US20230360411A1 (en) | Cherry picking and classifying method and device based on machine vision | |
| Liu et al. | Development of a machine vision algorithm for recognition of peach fruit in a natural scene | |
| CN110866547B (en) | Automatic classification system and method for traditional Chinese medicine decoction pieces based on multiple features and random forests | |
| Rahman et al. | Identification of mature grape bunches using image processing and computational intelligence methods | |
| CN113252584A (en) | Crop growth detection method and system based on 5G transmission | |
| CN111523503A (en) | Apple target detection method based on improved SSD algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20100317 |