CN107316043A - A kind of stacking mushroom method for quickly identifying of picking robot - Google Patents
A kind of stacking mushroom method for quickly identifying of picking robot Download PDFInfo
- Publication number
- CN107316043A CN107316043A CN201710544446.4A CN201710544446A CN107316043A CN 107316043 A CN107316043 A CN 107316043A CN 201710544446 A CN201710544446 A CN 201710544446A CN 107316043 A CN107316043 A CN 107316043A
- Authority
- CN
- China
- Prior art keywords
- mrow
- mushroom
- region
- image
- stacking
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of stacking mushroom method for quickly identifying of picking robot, by CCD camera continuous acquisition mushroom image, image segmentation computing is carried out to the image collected, background and noise is removed;Maximum mushroom region in computing screening target image is screened by region area, computing then is fitted using minimum circumscribed circle method to the region, border circular areas is obtained;By calculating each point in circle mushroom region home position is determined to the maximum of the minimum range of contour edge;The matching template of target mushroom is intercepted according to home position.Finally, match cognization is carried out to mushroom using quick normalized crosscorrelation matching algorithm.The inventive method accurately can quickly recognize stacking fruit, and picking robot crawl fruit registration is high, and run time is short, and real-time is higher.
Description
Technical field
The present invention relates to harvest machinery field, and in particular to the image recognition algorithm of fruit and vegetable picking robot, particularly pair
The method for quickly identifying of mushroom near-spherical fruit.
Background technology
In China, mushroom-cultivating has formed factorial praluction, and harvesting mushroom is carried out in dark moist mushroom house,
Working environment is poor.The application of picking robot can largely reduce labour, improve operating efficiency, improve working environment.In recent years
Come, various countries carry out the research to all kinds of fruit and vegetable picking robots in succession, however, due to the unstructured feature of mushroom growth, existing
The factor of many influence mushroom identifications, therefore, the quick identification of mushroom turns into the research emphasis of picking robot.
Mushroom stacking is more, growth posture is uncontrollable, images match difficulty all gives terminal executor of picking robot
Identification and positioning belt come difficult.In addition, present mushroom identification technology is difficult to differentiate between ripe mushroom and immature mushroom, not damaged
Mushroom and there is damage mushroom, it is difficult to realize the selectivity harvesting of mushroom, therefore, the quick identification of mushroom and orientation problem have been passed through
As one of picking robot key issue urgently to be resolved hurrily.
The content of the invention
The deficiency existed for prior art, fruit is stacked it is an object of the invention to provide one kind for near-sphericals such as mushrooms
Quick recognition positioning method, solve due to mushroom grow naturally cause stacking influence robot quickly recognize positioning the problem of.
Its method is simple, versatility is good, can accurately improve the picking rate of picking robot.
The present invention is achieved through the following technical solutions object above.
A kind of stacking mushroom method for quickly identifying of picking robot of the present invention, step is as follows:
Step 1:Stack mushroom image collection:Using ccd video camera continuous acquisition image;
Step 2:Stack mushroom segmentation:Dynamic threshold segmentation is carried out to the image collected, background is removed, and use mathematics
Morphological method carries out perfect, removal noise and hole and separating piled mushroom to the region after segmentation;
Step 3:Maximum mushroom extracted region:Region area screening is carried out to image, maximum mushroom region is obtained;
Step 4:Determine the central coordinate of circle of target mushroom:Circular fit is carried out to the target mushroom region after screening, and really
Central coordinate of circle after fixed fitting;
Step 5:Match cognization and positioning:According to central coordinate of circle, using the matching of quick normalized crosscorrelation to stacking mushroom
Carry out quick identification and positioning.
Further, used in the step 2 Mathematical Morphology Method the region after segmentation is carried out perfect process for:
Step 2.1, erosion operation is carried out to the region after segmentation using the circular configuration of a certain size radius, tried to achieve afterwards
The largest connected region in region, ambient noise and stacking portion are removed;
Step 2.2, dilation operation, filling region hole are carried out to region.
Further, the step 4 improves process and is:
Step 4.1:Target mushroom zoned circular is fitted:Computing is fitted to target area using minimum circumscribed circle method,
Final goal region is covered by border circular areas;
Step 4.2:It is determined that fitting border circular areas central coordinate of circle:By calculating the interior each point of circle to contour edge minimum range
Maximum finds out fitting border circular areas home position and with coordinatograph.
Further, the step 5 improves process and is:
In formula, (i, j) is subgraph FI, jCoordinate of the upper left corner in image F, referred to as reference point;R (i, j) is cross-correlation
Coefficient.
A kind of advantage of stacking mushroom method for quickly identifying of picking robot of the present invention is:For near-sphericals such as mushrooms
Fruit cultivates situation naturally, and the inventive method accurately can quickly recognize stacking fruit, and picking robot crawl fruit weight
Right height, run time is short, and real-time is higher.
Brief description of the drawings
Fig. 1 is to stack the quick identification process figure of mushroom;
Fig. 2 is to stack the image after mushroom image segmentation and morphology operations, wherein, Fig. 2 a are original image, Fig. 2 b
For the image after morphological erosion, Fig. 2 c are the image after morphological dilations;
Fig. 3 is to stack maximum mushroom extraction figure.
Fig. 4 is maximum mushroom circular fit figure.
Fig. 5 is the matching template figure extracted.
Fig. 6 is quick normalized crosscorrelation matching result figure.
Embodiment
Patent particular flow sheet of the present invention as shown in figure 1, the specific steps of patent of the present invention are made below in conjunction with the accompanying drawings into
The explanation of one step.
1st, mushroom image collection is stacked
The present invention uses CCD camera continuous acquisition image, and frequency acquisition is 10 frames/second.
2nd, mushroom segmentation is stacked
This patent, first using splitting based on grey level histogram dynamic threshold, separates mushroom and background area to image
Domain, then removes hole, noise etc. to the image after segmentation using Mathematical Morphology Method.Specific method is to use radius first
Erosion operation is carried out to image for 150 circular configuration element, the largest connected region in region is tried to achieve afterwards, by ambient noise and
Stacking portion is removed;Radius is finally used to carry out dilation operation, filling region hole for 130 circular configuration element to region.
It is as shown in Figure 2 that effect is improved in image segmentation.
3rd, maximum mushroom extracted region
By being screened to segmenting the image deployment area area dealt with problems arising from an accident, maximum mushroom region is filtered out.Specific screening fortune
Calculation method is that the point for carrying out certain specification size to target area first is marked so that whole target area, which is put, to be filled up, then
The number put in each subregion is calculated, the point region of quantity at most is maximum mushroom region.Maximum mushroom extracted region figure is such as
Shown in Fig. 3.
4th, the central coordinate of circle of target mushroom is determined
Computing is fitted to target area using minimum circumscribed circle method first, target area is covered by border circular areas;
Then look for the position that each point in circle finds out the center of circle to the maximum of the minimum range of contour edge.Specific method is to circle
Point in domain scan using multi-direction from profile into circle, the minimum range of calculating each point to contour edge, finally these away from
From maximum is selected in value, the position of the maximum corresponding points is center of circle position.Result figure is as shown in Figure 4.
5th, target mushroom template is extracted
The template that fitted area and central coordinate of circle interception according to being determined in 4 matches.Result figure is as shown in Figure 5.
6th, match cognization and positioning
According to central coordinate of circle, quick identification and positioning are carried out to stacking mushroom using the matching of quick normalized crosscorrelation.
It is as shown in Figure 6 with result figure.
The algorithm steps of normalized crosscorrelation matching are as follows:
If image F (pixel M × N) to be matched and template image T (pixel m × n).The definition of normalizated correlation coefficient is:
Prototype drawing T is placed on to search on sketch map F and translated, that block search figure under template covering is called FI, j。
In formula:(i, j) is subgraph FI, jCoordinate of the upper left corner in image F, referred to as reference point.The item of molecular position is
The cross-correlation of subgraph and template image, changes with the change of (i, j);The Section 1 of denominator represents subgraph FI, jGross energy,
What the Section 2 of denominator was represented is template T gross energy.
R (i, j) scope is between (0,1), and R is the metric of similitude, and the value is bigger, illustrates matching template phase knowledge and magnanimity
It is higher.As template T and covering subgraph match, there is maximum in R, when template and just the same covering subgraph, cross correlation
Number R (i, j)=1.After whole search graph F is completed, cross-correlation coefficient R is found outmax(i, j), its corresponding FI, jSeek to what is looked for
Match target.
Claims (5)
1. the stacking mushroom method for quickly identifying of a kind of picking robot, it is characterised in that comprise the following steps:
Step 1:Stack mushroom image collection:Using ccd video camera continuous acquisition image;
Step 2:Stack mushroom segmentation:Dynamic threshold segmentation is carried out to the image collected, background is removed, and use Mathematical Morphology
Method carries out perfect, removal noise and hole and separating piled mushroom to the region after segmentation;
Step 3:Maximum mushroom extracted region:Region area screening is carried out to image, maximum mushroom region is obtained;
Step 4:Determine the central coordinate of circle of target mushroom:Circular fit is carried out to the target mushroom region after screening, and determines to intend
Central coordinate of circle after conjunction;
Step 5:Match cognization and positioning:According to central coordinate of circle, carried out using the matching of quick normalized crosscorrelation to stacking mushroom
Quick identification and positioning.
2. the stacking mushroom method for quickly identifying of picking robot according to claim 1, it is characterised in that the step
Used in 2 Mathematical Morphology Method to after segmentation image carry out improve process for:
Step 2.1, use radius to carry out erosion operation to the region after segmentation for 150 circular configuration, region is tried to achieve afterwards
Largest connected region, ambient noise and stacking portion are removed;
Step 2.2, radius is used to carry out dilation operation, filling region hole to region for 130 circular configuration.
3. the stacking mushroom method for quickly identifying of picking robot according to claim 1, it is characterised in that the step
3 specific methods are:
The point for carrying out certain specification size to target area first is marked so that whole target area, which is put, to be filled up, and is then calculated
The number put in each subregion, the point region of quantity at most is maximum mushroom region.
4. the stacking mushroom method for quickly identifying of picking robot according to claim 1, it is characterised in that the step
4, which improve process, is:
Step 4.1:Target mushroom zoned circular is fitted:Computing is fitted to target area using minimum circumscribed circle method, finally
Target area is covered by border circular areas;
Step 4.2:It is determined that fitting border circular areas central coordinate of circle:By calculating the interior each point of circle to the very big of contour edge minimum range
Value finds out fitting border circular areas home position and with coordinatograph.
5. the stacking mushroom method for quickly identifying of picking robot according to claim 1, it is characterised in that the step
5, which improve process, is:
According to central coordinate of circle, quick identification and positioning are carried out to stacking mushroom using the matching of quick normalized crosscorrelation.
The algorithm steps of normalized crosscorrelation matching are as follows:
If image F (pixel M × N) to be matched and template image T (pixel m × n).The definition of normalizated correlation coefficient is:
<mrow>
<mi>R</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>M</mi>
</msubsup>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>n</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</msubsup>
<mo>&lsqb;</mo>
<msub>
<mi>F</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>&times;</mo>
<mi>T</mi>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
</mrow>
<mrow>
<msqrt>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>M</mi>
</msubsup>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>n</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</msubsup>
<msup>
<mrow>
<mo>&lsqb;</mo>
<msub>
<mi>F</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
<msqrt>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>M</mi>
</msubsup>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>n</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</msubsup>
<msup>
<mrow>
<mo>&lsqb;</mo>
<mi>T</mi>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
</mrow>
</mfrac>
</mrow>
Prototype drawing T is placed on to search on sketch map F and translated, that block search figure under template covering is called FI, j。
In formula:(i, j) is subgraph FI, jCoordinate of the upper left corner in image F, referred to as reference point.The item of molecular position is subgraph
The cross-correlation of picture and template image, changes with the change of (i, j);The Section 1 of denominator represents subgraph F gross energy, denominator
What Section 2 was represented is template T gross energy.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710544446.4A CN107316043A (en) | 2017-07-04 | 2017-07-04 | A kind of stacking mushroom method for quickly identifying of picking robot |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710544446.4A CN107316043A (en) | 2017-07-04 | 2017-07-04 | A kind of stacking mushroom method for quickly identifying of picking robot |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107316043A true CN107316043A (en) | 2017-11-03 |
Family
ID=60180122
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710544446.4A Pending CN107316043A (en) | 2017-07-04 | 2017-07-04 | A kind of stacking mushroom method for quickly identifying of picking robot |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107316043A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108171693A (en) * | 2017-12-27 | 2018-06-15 | 合肥市雅视智能科技有限公司 | A kind of method of automatic detection poor quality mushroom |
CN108805875A (en) * | 2018-04-17 | 2018-11-13 | 上海大学 | White mushroom position identifies and pushes over the method for selecting in direction |
CN110059663A (en) * | 2019-04-24 | 2019-07-26 | 扬州龙科信息科技有限公司 | A kind of visual identity of adhesion mushroom and measurement method based on characteristic point detection |
CN110084799A (en) * | 2019-04-24 | 2019-08-02 | 扬州龙科信息科技有限公司 | A kind of area brown mushroom Gai San vision On-line Measuring Method based on depth camera |
CN111445482A (en) * | 2020-03-24 | 2020-07-24 | 上海第二工业大学 | Segmentation and identification method of overlapped agaricus bisporus |
CN112183230A (en) * | 2020-09-09 | 2021-01-05 | 上海大学 | Identification and central point positioning method for pears in natural pear orchard environment |
CN114347044A (en) * | 2022-03-17 | 2022-04-15 | 季华实验室 | Pepper picking method and device, electronic equipment and storage medium |
CN116310806A (en) * | 2023-02-28 | 2023-06-23 | 北京理工大学珠海学院 | Intelligent agriculture integrated management system and method based on image recognition |
-
2017
- 2017-07-04 CN CN201710544446.4A patent/CN107316043A/en active Pending
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108171693A (en) * | 2017-12-27 | 2018-06-15 | 合肥市雅视智能科技有限公司 | A kind of method of automatic detection poor quality mushroom |
CN108805875A (en) * | 2018-04-17 | 2018-11-13 | 上海大学 | White mushroom position identifies and pushes over the method for selecting in direction |
CN108805875B (en) * | 2018-04-17 | 2022-07-08 | 上海大学 | Method for identifying position of agaricus bisporus and selecting pushing direction |
CN110059663A (en) * | 2019-04-24 | 2019-07-26 | 扬州龙科信息科技有限公司 | A kind of visual identity of adhesion mushroom and measurement method based on characteristic point detection |
CN110084799A (en) * | 2019-04-24 | 2019-08-02 | 扬州龙科信息科技有限公司 | A kind of area brown mushroom Gai San vision On-line Measuring Method based on depth camera |
CN111445482A (en) * | 2020-03-24 | 2020-07-24 | 上海第二工业大学 | Segmentation and identification method of overlapped agaricus bisporus |
CN111445482B (en) * | 2020-03-24 | 2023-03-28 | 上海第二工业大学 | Segmentation and identification method of overlapped agaricus bisporus |
CN112183230A (en) * | 2020-09-09 | 2021-01-05 | 上海大学 | Identification and central point positioning method for pears in natural pear orchard environment |
CN114347044A (en) * | 2022-03-17 | 2022-04-15 | 季华实验室 | Pepper picking method and device, electronic equipment and storage medium |
CN114347044B (en) * | 2022-03-17 | 2022-06-17 | 季华实验室 | Pepper picking method and device, electronic equipment and storage medium |
CN116310806A (en) * | 2023-02-28 | 2023-06-23 | 北京理工大学珠海学院 | Intelligent agriculture integrated management system and method based on image recognition |
CN116310806B (en) * | 2023-02-28 | 2023-08-29 | 北京理工大学珠海学院 | Intelligent agriculture integrated management system and method based on image recognition |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107316043A (en) | A kind of stacking mushroom method for quickly identifying of picking robot | |
CN107945192B (en) | Tray carton pile type real-time detection method | |
CN113658132B (en) | Computer vision-based structural part weld joint detection method | |
Fernandez et al. | Automatic detection of field-grown cucumbers for robotic harvesting | |
CN104636722B (en) | A kind of overlapping fruit quick Tracking Recognition method of picking robot | |
CN103310218B (en) | A kind of overlap blocks fruit precise recognition method | |
CN111844101B (en) | Multi-finger dexterous hand sorting planning method | |
CN111666883B (en) | Grape picking robot target identification and fruit stalk clamping and cutting point positioning method | |
CN112132153B (en) | Tomato fruit identification method and system based on clustering and morphological processing | |
CN107527343B (en) | A kind of agaricus bisporus stage division based on image procossing | |
CN103295018A (en) | Method for precisely recognizing fruits covered by branches and leaves | |
CN106875412B (en) | Segmentation positioning method for two overlapped fruits | |
CN112990103B (en) | String mining secondary positioning method based on machine vision | |
CN107808141A (en) | A kind of electric transmission line isolator explosion recognition methods based on deep learning | |
CN104091175B (en) | A kind of insect automatic distinguishing method for image based on Kinect depth information acquiring technology | |
JP6288230B2 (en) | Object division method and apparatus | |
CN111046782B (en) | Quick fruit identification method for apple picking robot | |
CN109684941A (en) | One kind picking region partitioning method based on MATLAB image procossing litchi fruits | |
CN111445482A (en) | Segmentation and identification method of overlapped agaricus bisporus | |
CN112380926A (en) | Weeding path planning system of field weeding robot | |
Yoshida et al. | A tomato recognition method for harvesting with robots using point clouds | |
Liu et al. | SE-Mask R-CNN: An improved Mask R-CNN for apple detection and segmentation | |
CN113450402B (en) | Navigation center line extraction method for vegetable greenhouse inspection robot | |
WO2021228194A1 (en) | Cable detection method, robot and storage device | |
Rahman et al. | Identification of mature grape bunches using image processing and computational intelligence methods |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171103 |
|
WD01 | Invention patent application deemed withdrawn after publication | ||
DD01 | Delivery of document by public notice |
Addressee: Li Minghang Document name: Deemed as a notice of withdrawal |
|
DD01 | Delivery of document by public notice |