CN103528967A - Hyperspectral image based overripe Lonicera edulis fruit identification method - Google Patents
Hyperspectral image based overripe Lonicera edulis fruit identification method Download PDFInfo
- Publication number
- CN103528967A CN103528967A CN201310463530.5A CN201310463530A CN103528967A CN 103528967 A CN103528967 A CN 103528967A CN 201310463530 A CN201310463530 A CN 201310463530A CN 103528967 A CN103528967 A CN 103528967A
- Authority
- CN
- China
- Prior art keywords
- fruit
- pixel
- overdone
- wave band
- image
- 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
Images
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention relates to a hyperspectral image based overripe Lonicera edulis fruit identification method. The method comprises the following steps: 1, acquiring a hyperspectral image of a Lonicera edulis fruit; 2, sampling fruit and background pixels in the hyperspectral image, analyzing and providing an image background removal function model; 3, removing the image background according to the functional model established in step 2; 4, adopting a median filtering, morphological filtering, spatial processing and threshold determining method to remove noises and determine the position of the fruit; 5, sampling pixels of an overripe fruit and a ripe fruit, selecting a most discriminative waveband through a stepwise forward variable selection method, and adopting a linear discriminant analysis method to establish and discriminate the function models of the overripe fruit pixel and the ripe fruit pixel; 6, classifying all fruit pixels according to the discrimination models established in step 5, and respectively marking; and 7, adopting a majority principle to classify all the fruits. The identification method utilizes the spectral information of objectives and also utilizes the spatial information of fruits, so the discrimination rate is improved.
Description
Technical field
The present invention relates to infrared spectrum identification field, be specifically related to a kind of overdone indigo fruit fruit recognition methods based on high spectrum image.
Background technology
Indigo fruit is perennial machaka, its fruits nutrition is worth very high, contain the multiple elements such as ash content, protein, fat, tannin, pectin, volatile acid, vitamin and phosphorus, particularly ascorbic content is higher, and it also contains 7 seed amino acids and various trace elements that the mankind must obtain from food.Fresh fruit can be eaten raw, and the good raw material of brewed fruit wine and beverage, is rare natural colouring matter product especially.Its berry also can be used as medicine, and has antipyretic and antidote functions.In the indigo fruit fruit of gathering in the crops by mechanical-vibration type picker, comprise underdone redness, cyan fruit, just ripe and overdone mazarine fruit.Wherein, underdone redness, cyan fruit can judge by common RGB image processing method; And overdone fruit is identical with the fruit surface color of proper mature (ripe), cannot adopt said method to distinguish.But overdone fruit is partially soft, in transportation, very easily breakage causes pulp to leak outside and affects other normal fruits, reduces its marketable value.Conventional method is the artificial hardness with finger judgement fruit at present, and partially soft fruit is picked out, and workload is large and efficiency is low.
Hyper-spectral image technique is that imaging technique is combined with spectrographic detection technology, in the space characteristics imaging to target, each space pixel is covered to carry out continuous spectrum through tens of dispersion formation and even a hundreds of narrow wave band.High-spectrum image set image information and spectral information are.Image information can reflect the external sort features such as the size, shape, defect of imaging object, because heterogeneity is also different to spectral absorption, at certain specific wavelength hypograph, certain defect is had to reflection more significantly, and spectral information can fully reflect the physical arrangement of sample interior, the difference of chemical composition.These features have determined the unique advantage of hyper-spectral image technique in the context of detection of agricultural product inside quality.
Summary of the invention
Analysis based on above, the object of the invention is to utilize high spectrum image, for the identification of overdone indigo fruit fruit provides a kind of brand-new method.
In order to solve the problems of the technologies described above, the invention provides a kind of recognition methods of the overdone indigo fruit fruit based on high spectrum image, comprise step:
S1: indigo fruit fruit to be identified is inserted to high light spectrum image-forming device, gather high spectrum image;
S2: respectively fruit pixel and background pixel in high spectrum image are sampled, analyze its spectral signature, propose to remove the function model of image background;
S3: remove image background according to the function model of setting up in S2;
S4: adopt the method for medium filtering, morphologic filtering, spatial manipulation and threshold decision to remove noise, thereby determine the position of fruit;
S5: respectively overdone fruit pixel and proper mature (ripe) fruit pixel are sampled, utilize progressively variable forward to select the method for (Forward stepwise variable selection) to analyze spectral information, thereby optimize the wave band most with identification, then adopt linear discriminant analysis (Linear discriminant analysis) method to set up the function model of the overdone fruit pixel of differentiation and proper mature (ripe) fruit pixel;
S6: the discrimination model of setting up according to S5 calculates and identification and classification each pixel of each fruit, and be labeled as respectively overdone fruit pixel or proper mature (ripe) fruit pixel;
S7: adopt majority principle to classify to each fruit, when differentiation surpasses the pixel count of a fruit 50% for overdone pixel count, this fruit is judged to overdone fruit, otherwise be proper mature (ripe) fruit.
Whole samples of every class fruit are divided into two parts at random, and a part of sample is for modeling, and another part sample is for test;
If high spectrum image comprises p continuous wave band, arbitrary pixel is at the brightness value I of any wave band q
q;
Wherein, the acquisition methods of the function model of the removal image background of described step S2 is:
Respectively fruit pixel and background pixel in high spectrum image are sampled, the average brightness value of getting each comfortable fruit pixel and background pixel to each wave band, obtain the spectral pattern of fruit pixel and background pixel, find respectively the brightness maximal value place wave band of these two curves, if the maximum brightness value place wave band of the spectral pattern of fruit pixel is m, the maximum brightness value place wave band of the spectral pattern of background pixel is n, and the model that proposes thus removal image background is:
V
B=I
m/I
n
By experimental study, determine a threshold value V, work as V
bduring > V, this pixel belongs to fruit pixel, otherwise this pixel belongs to background pixel, thereby obtains the preliminary bianry image of removing background;
Wherein, the threshold decision in described step S4 is:
According to the resolution of concrete imaging device and with the distance of fruit, carry out analysis of experiments, determine a threshold value A, in image, the pixel count in certain region is greater than A, is judged to be fruit region, otherwise, be judged to be noise region and process;
Wherein, in described step S5, the method for the function model of the foundation overdone fruit pixel of differentiation and proper mature (ripe) fruit pixel is:
S51: respectively overdone fruit pixel and proper mature (ripe) fruit pixel are sampled;
S52: utilize progressively variable forward to select the method for (Forward stepwise variable selection) to analyze the spectral information of all p wave band, thereby optimize k the wave band x most with identification
1, x
2..., x
k;
S53: utilize the spectral information of an above-mentioned k wave band, adopt linear discriminant analysis (Linear discriminant analysis) method to set up the discriminant function model of the overdone fruit pixel of differentiation and proper mature (ripe) fruit pixel: y=b
0+ b
1i
x1-b
2i
x2+ ...+b
ki
xk, y is discriminant value, b
0constant term, b
jbe j (j=1,2 ..., k) discriminant coefficient of individual wave band.
The method that the present invention proposes, has not only utilized the spectral information of object, has also utilized the spatial information of fruit, and this two category information is combined, and improves differentiation rate.Judge that with existing finger prosthesis thereby the hardness of fruit determines that its whether overdone method compares, the method that the present invention proposes, to not injury of fruit, for Non-Destructive Testing, and preparation rate is high, there is scientific basis, and can utilize the selected characteristic wave bands of the method to develop corresponding screening installation, there is very strong technical advantage.
Accompanying drawing explanation
Fig. 1 is the process flow diagram that the present invention is based on the overdone indigo fruit fruit recognition methods of high spectrum image;
Fig. 2 is the spectral pattern of fruit and background in high spectrum image;
Fig. 3 is the overdone indigo fruit fruit identifying based on high spectrum image; Wherein (a) is the RGB image of indigo fruit fruit, and odd number is classified overdone fruit as, and even number is classified proper mature (ripe) fruit as; (b) gray-scale map of locating at wave band 35 (751nm) for the high spectrum image of fruit shown in (a); (c) for removing the image after background according to the function model of setting up in step S2, white represents to be identified as fruit pixel, and black represents to be identified as background pixel; (d) for adopting the method for medium filtering, morphologic filtering, spatial manipulation and threshold decision to remove the fruit location drawing picture after noise; (e) gray-scale map of locating at wave band 35 (751nm) for the fruit high spectrum image extracting according to fruit position shown in (d); (f) for each fruit pixel being carried out to the result after identification and classification according to the identification and classification model of the overdone fruit pixel of step S5 foundation and proper mature (ripe) fruit pixel, wherein gray pixels is overdone fruit pixel, and white pixel is proper mature (ripe) fruit pixel; (g) be the fruit classification definite according to the majority principle of step S7, grey represents to be identified as overdone fruit, and white represents to be identified as proper mature (ripe) fruit; (h) be actual fruit classification, it is overdone fruit that grey represents actual, and white represents that reality is proper mature (ripe) fruit.
Embodiment
Below in conjunction with drawings and Examples, the specific embodiment of the present invention is described in further detail.
As shown in Figure 1, the overdone indigo fruit fruit recognition methods based on high spectrum image of the present invention comprises:
Step S1, inserts the indigo fruit fruit of collection in high light spectrum image-forming device, gathers high spectrum image, the spectral wavelength scope of this example is 369-1042nm, comprise 60 wave bands, the scope of each wave band is 11nm, and the computing formula of the centre wavelength of each wave band is:
w
l=0.000324(8w
n+1)
2+1.273(8w
n+1)+367.7
W wherein
lcentered by wavelength, w
nfor ripple segment number, the value of n is 0,1,2 ..., 59;
Step S2: respectively fruit pixel and background pixel in high spectrum image are sampled, obtain spectral pattern as shown in Figure 2, wherein fruit is obtained maximal value when wave band 35 (751nm), background obtains maximal value when wave band 28 (671nm), and the model that therefore proposes removal image background is:
V
B=I
35/I
28
I wherein
35the brightness value that pixel is located at wave band 35 (751nm), I
28it is the brightness value that pixel is located at wave band 28 (671nm);
Step S3: according to the function model of setting up in S2, get the brightness value that each pixel in image is located at wave band 35 (751nm) and wave band 28 (671nm), compare rear acquisition V
b, the V obtaining with experimental study
b=1.0 is threshold value, works as V
bduring > 1.0, this pixel belongs to fruit pixel, otherwise, this pixel belongs to background pixel, thereby obtains the preliminary bianry image of removing background, as shown in Fig. 3 (b), white represents to be identified as fruit pixel, and black represents to be identified as background pixel;
Step S4, for above-mentioned bianry image, the operator of employing 9 * 9 carries out corroding operation after medium filtering, the pixel region that the method that process application space is again about to connect carries out mark by identical numeral, unconnected pixel is carried out mark by different numerals, thereby obtain number of regions in image and the pixel count in each region, 400 pixels that the experimental study of finally take is determined are threshold value, pixel count is less than to 400 region as noise and removes, determine thus fruit quantity and the position in image, and the pixel count of each fruit, white portion as shown in Fig. 3 (d) is the position of fruit,
Step S5, respectively overdone fruit pixel and proper mature (ripe) fruit pixel are sampled, utilize the method that progressively variable is selected forward to analyze spectral information, thereby optimizing the wave band most with identification is wave band 35 (751nm) and wave band 5 (420nm), then adopt linear discriminant analysis method to set up to differentiate overdone fruit pixel and proper mature (ripe) really the function model of pixel be
y=2.317×I
35-4.057×I
5-5.593
I wherein
35the brightness value that pixel is located at wave band 35 (751nm), I
5it is the brightness value that pixel is located at wave band 5 (420nm);
Step S6: each fruit pixel of determining for S4, extract the brightness value that it is located at wave band 35 (751nm) and wave band 5 (420nm), the definite discriminative model of application S5 is classified to each fruit pixel, when y > 0, this pixel is judged as overdone fruit pixel, otherwise, this pixel is judged as proper mature (ripe) fruit pixel, as shown in Fig. 3 (f), wherein gray pixels is overdone fruit pixel, and white pixel is proper mature (ripe) fruit pixel;
Step S7: for each fruit, adopt majority principle finally to classify to it, when the overdone fruit pixel count of a fruit surpass this fruit total pixel number 50% time, this fruit is judged as overdone fruit, otherwise this fruit is judged as proper mature (ripe) fruit, as shown in Fig. 3 (g), grey represents to be identified as overdone fruit, and white represents to be identified as proper mature (ripe) fruit.
Claims (4)
1. the overdone indigo fruit fruit recognition methods based on high spectrum image, is characterized in that, comprises step:
S1: indigo fruit fruit to be identified is inserted to high light spectrum image-forming device, gather high spectrum image;
S2: respectively fruit pixel and background pixel in high spectrum image are sampled, analyze its spectral signature, propose to remove the function model of image background;
S3: remove image background according to the function model of setting up in S2;
S4: adopt the method for medium filtering, morphologic filtering, spatial manipulation and threshold decision to remove noise, thereby determine the position of fruit;
S5: respectively overdone fruit pixel and proper mature (ripe) fruit pixel are sampled, utilize the method that progressively variable is selected forward to analyze spectral information, thereby optimize the wave band most with identification, then adopt linear discriminant analysis method to set up the function model of the overdone fruit pixel of differentiation and proper mature (ripe) fruit pixel;
S6: the discrimination model of setting up according to S5 calculates and identification and classification each pixel of each fruit, and be labeled as respectively overdone fruit pixel or proper mature (ripe) fruit pixel;
S7: adopt majority principle to classify to each fruit, when differentiation surpasses the pixel count of a fruit 50% for overdone pixel count, this fruit is judged to overdone fruit, otherwise be proper mature (ripe) fruit.
2. a kind of overdone indigo fruit fruit recognition methods based on high spectrum image as claimed in claim 1, is characterized in that:
If high spectrum image comprises p continuous wave band, arbitrary pixel is at the brightness value I of any wave band q
q;
Wherein, the acquisition methods of the function model of the removal image background of described step S2 is:
Respectively fruit pixel and background pixel in high spectrum image are sampled, the average brightness value of getting each comfortable fruit pixel and background pixel to each wave band, obtain the spectral pattern of fruit pixel and background pixel, find respectively the brightness maximal value place wave band of these two curves, if the maximum brightness value place wave band of the spectral pattern of fruit pixel is m, the maximum brightness value place wave band of the spectral pattern of background pixel is n, and the model that proposes thus removal image background is:
V
B=I
m/I
n
By experimental study, determine a threshold value V, work as V
bduring > V, this pixel belongs to fruit pixel, otherwise this pixel belongs to background pixel, thereby obtains the preliminary bianry image of removing background.
3. a kind of overdone indigo fruit fruit recognition methods based on high spectrum image as claimed in claim 1, is characterized in that, the threshold decision in wherein said step S4 is:
According to the resolution of concrete imaging device and with the distance of fruit, carry out analysis of experiments, determine a threshold value A, in image, the pixel count in certain region is greater than A, is judged to be fruit region, otherwise, be judged to be noise and process.
4. a kind of overdone indigo fruit fruit recognition methods based on high spectrum image as claimed in claim 1, is characterized in that, the method for setting up the function model of differentiating overdone fruit pixel and proper mature (ripe) fruit pixel in wherein said step S5 is:
S51: respectively overdone fruit pixel and proper mature (ripe) fruit pixel are sampled;
S52: utilize progressively the method that variable is selected forward to analyze the spectral information of all p wave band, thereby optimize k the wave band x most with identification
1, x
2..., x
k;
S53: utilize the spectral information of an above-mentioned k wave band, adopt linear discriminant analysis method to set up the discriminant function model of the overdone fruit pixel of differentiation and proper mature (ripe) fruit pixel: y=b
0+ b
1i
x1-b
2i
x2+ ...+b
ki
xk, y is discriminant value, b
0constant term, b
jbe j (j=1,2 ..., k) discriminant coefficient of individual wave band.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310463530.5A CN103528967B (en) | 2013-09-25 | 2013-09-25 | Overdone indigo fruit fruit identification method based on high spectrum image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310463530.5A CN103528967B (en) | 2013-09-25 | 2013-09-25 | Overdone indigo fruit fruit identification method based on high spectrum image |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103528967A true CN103528967A (en) | 2014-01-22 |
CN103528967B CN103528967B (en) | 2016-08-31 |
Family
ID=49931154
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310463530.5A Expired - Fee Related CN103528967B (en) | 2013-09-25 | 2013-09-25 | Overdone indigo fruit fruit identification method based on high spectrum image |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103528967B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103822879A (en) * | 2014-02-24 | 2014-05-28 | 西北农林科技大学 | Nondestructive detection method of swelled kiwi fruits based on hyperspectral imaging technology |
CN106644958A (en) * | 2016-11-14 | 2017-05-10 | 浙江大学 | Method for imaging spatial distribution of pectin content in peach fruit |
CN106871569A (en) * | 2016-12-30 | 2017-06-20 | 青岛海尔智能技术研发有限公司 | The category identification method and refrigerator of refrigerator inside food |
CN107505325A (en) * | 2017-08-18 | 2017-12-22 | 西北农林科技大学 | The comprehensive quality detecting method of winter-jujube fruit |
CN113252585A (en) * | 2021-04-21 | 2021-08-13 | 深圳市海谱纳米光学科技有限公司 | Method and device for judging gold surface coating based on hyperspectral image |
CN113959961A (en) * | 2021-12-22 | 2022-01-21 | 广东省农业科学院动物科学研究所 | Hyperspectral image-based tannin additive anti-counterfeiting detection method and system |
CN117288692A (en) * | 2023-11-23 | 2023-12-26 | 四川轻化工大学 | Method for detecting tannin content in brewing grains |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1868610A (en) * | 2006-04-29 | 2006-11-29 | 江西农业大学 | Small-size movable fruit and birds, beasts and eggs intelligence grading plant |
CN1995987A (en) * | 2007-02-08 | 2007-07-11 | 江苏大学 | Non-destructive detection method and device for agricultural and animal products based on hyperspectral image technology |
CN102661920A (en) * | 2012-04-11 | 2012-09-12 | 东北林业大学 | Device and method for classifying and identifying weeds |
CN102706813A (en) * | 2012-05-31 | 2012-10-03 | 北京林业大学 | Poa pratensis variety identification method based on hyper-spectral image |
CN103033512A (en) * | 2012-07-24 | 2013-04-10 | 南京农业大学 | Device and method for recognizing hatching egg incubation based on hyperspectrum |
CN103111427A (en) * | 2013-01-10 | 2013-05-22 | 上海交通大学 | Fruit grader based on image processing |
-
2013
- 2013-09-25 CN CN201310463530.5A patent/CN103528967B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1868610A (en) * | 2006-04-29 | 2006-11-29 | 江西农业大学 | Small-size movable fruit and birds, beasts and eggs intelligence grading plant |
CN1995987A (en) * | 2007-02-08 | 2007-07-11 | 江苏大学 | Non-destructive detection method and device for agricultural and animal products based on hyperspectral image technology |
CN102661920A (en) * | 2012-04-11 | 2012-09-12 | 东北林业大学 | Device and method for classifying and identifying weeds |
CN102706813A (en) * | 2012-05-31 | 2012-10-03 | 北京林业大学 | Poa pratensis variety identification method based on hyper-spectral image |
CN103033512A (en) * | 2012-07-24 | 2013-04-10 | 南京农业大学 | Device and method for recognizing hatching egg incubation based on hyperspectrum |
CN103111427A (en) * | 2013-01-10 | 2013-05-22 | 上海交通大学 | Fruit grader based on image processing |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103822879A (en) * | 2014-02-24 | 2014-05-28 | 西北农林科技大学 | Nondestructive detection method of swelled kiwi fruits based on hyperspectral imaging technology |
CN103822879B (en) * | 2014-02-24 | 2016-08-17 | 西北农林科技大学 | A kind of Fructus actinidiae chinensis based on high light spectrum image-forming technology expand fruit lossless detection method |
CN106644958A (en) * | 2016-11-14 | 2017-05-10 | 浙江大学 | Method for imaging spatial distribution of pectin content in peach fruit |
CN106644958B (en) * | 2016-11-14 | 2019-04-05 | 浙江大学 | A kind of method of Peach fruits inside pectin content spatial distribution imaging |
CN106871569A (en) * | 2016-12-30 | 2017-06-20 | 青岛海尔智能技术研发有限公司 | The category identification method and refrigerator of refrigerator inside food |
CN107505325A (en) * | 2017-08-18 | 2017-12-22 | 西北农林科技大学 | The comprehensive quality detecting method of winter-jujube fruit |
CN107505325B (en) * | 2017-08-18 | 2023-04-25 | 西北农林科技大学 | Omnibearing quality detection method for winter jujube fruits |
CN113252585A (en) * | 2021-04-21 | 2021-08-13 | 深圳市海谱纳米光学科技有限公司 | Method and device for judging gold surface coating based on hyperspectral image |
CN113252585B (en) * | 2021-04-21 | 2024-02-09 | 深圳市海谱纳米光学科技有限公司 | Method and device for judging gold surface coating based on hyperspectral image |
CN113959961A (en) * | 2021-12-22 | 2022-01-21 | 广东省农业科学院动物科学研究所 | Hyperspectral image-based tannin additive anti-counterfeiting detection method and system |
CN117288692A (en) * | 2023-11-23 | 2023-12-26 | 四川轻化工大学 | Method for detecting tannin content in brewing grains |
CN117288692B (en) * | 2023-11-23 | 2024-04-02 | 四川轻化工大学 | Method for detecting tannin content in brewing grains |
Also Published As
Publication number | Publication date |
---|---|
CN103528967B (en) | 2016-08-31 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Bhargava et al. | Fruits and vegetables quality evaluation using computer vision: A review | |
Xing et al. | Combination of chemometric tools and image processing for bruise detection on apples | |
Bauriegel et al. | Early detection of Fusarium infection in wheat using hyper-spectral imaging | |
CN103528967A (en) | Hyperspectral image based overripe Lonicera edulis fruit identification method | |
Blasco et al. | Recognition and classification of external skin damage in citrus fruits using multispectral data and morphological features | |
CN109657653A (en) | A kind of wheat seed head blight recognition methods based on Imaging Hyperspectral Data | |
CN110763698B (en) | Hyperspectral citrus leaf disease identification method based on characteristic wavelength | |
Guzmán et al. | Infrared machine vision system for the automatic detection of olive fruit quality | |
Eissa et al. | Understanding color image processing by machine vision for biological materials | |
Abdullah et al. | Automated inspection system for colour and shape grading of starfruit (Averrhoa carambola L.) using machine vision sensor | |
CN107505325B (en) | Omnibearing quality detection method for winter jujube fruits | |
Zou et al. | Peanut maturity classification using hyperspectral imagery | |
CN105067532A (en) | Method for identifying early-stage disease spots of sclerotinia sclerotiorum and botrytis of rape | |
CN109883967A (en) | A kind of Eriocheir sinensis quality grade method of discrimination based on information fusion | |
CN110929787B (en) | Apple objective grading system based on image | |
Alfatni et al. | Recent methods and techniques of external grading systems for agricultural crops quality inspection-review | |
Yro et al. | Cocoa beans fermentation degree assessment for quality control using machine vision and multiclass svm classifier | |
CN116559111A (en) | Sorghum variety identification method based on hyperspectral imaging technology | |
Torres et al. | Setting up a methodology to distinguish between green oranges and leaves using hyperspectral imaging | |
CN113570538A (en) | Blade RGB image skewed distribution parameter information acquisition and analysis method | |
Janardhana et al. | Computer aided inspection system for food products using machine vision—a review | |
Jin et al. | Methodology for potatoes defects detection with computer vision | |
Ji et al. | Apple color automatic grading method based on machine vision | |
CN110118735B (en) | Hyperspectral imaging detection method and device for detecting male and female bergamot pears | |
Zhang et al. | Date quality evaluation using short-wave infrared imaging |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
CB03 | Change of inventor or designer information |
Inventor after: Fu Longsheng Inventor after: Sun Shipeng Inventor after: Feng Yali Inventor after: Peng Jun Inventor after: Li Rui Inventor before: Fu Longsheng Inventor before: Li Rui |
|
COR | Change of bibliographic data | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160831 Termination date: 20170925 |
|
CF01 | Termination of patent right due to non-payment of annual fee |