CN115375694A - Portable rice whole ear measuring method based on image recognition and application thereof - Google Patents
Portable rice whole ear measuring method based on image recognition and application thereof Download PDFInfo
- Publication number
- CN115375694A CN115375694A CN202211322966.8A CN202211322966A CN115375694A CN 115375694 A CN115375694 A CN 115375694A CN 202211322966 A CN202211322966 A CN 202211322966A CN 115375694 A CN115375694 A CN 115375694A
- Authority
- CN
- China
- Prior art keywords
- branch
- image
- rice
- branches
- primary
- 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
- 235000007164 Oryza sativa Nutrition 0.000 title claims abstract description 99
- 235000009566 rice Nutrition 0.000 title claims abstract description 99
- 238000000034 method Methods 0.000 title claims abstract description 57
- 240000007594 Oryza sativa Species 0.000 title 1
- 235000013339 cereals Nutrition 0.000 claims abstract description 114
- 241000209094 Oryza Species 0.000 claims abstract description 98
- 238000012937 correction Methods 0.000 claims abstract description 43
- 238000003702 image correction Methods 0.000 claims abstract description 8
- 238000004458 analytical method Methods 0.000 claims description 31
- 238000001514 detection method Methods 0.000 claims description 25
- 230000009466 transformation Effects 0.000 claims description 15
- 230000011218 segmentation Effects 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 13
- 238000000691 measurement method Methods 0.000 claims description 8
- 238000012216 screening Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 7
- 238000012805 post-processing Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 5
- 238000005286 illumination Methods 0.000 claims description 5
- 239000011159 matrix material Substances 0.000 claims description 5
- 238000003709 image segmentation Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 abstract description 8
- 210000005069 ears Anatomy 0.000 description 11
- 238000012545 processing Methods 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000009499 grossing Methods 0.000 description 3
- 238000009395 breeding Methods 0.000 description 2
- 230000001488 breeding effect Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000005260 corrosion Methods 0.000 description 2
- 230000007797 corrosion Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000000877 morphologic effect Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000016574 developmental growth Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000012010 growth Effects 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 235000016709 nutrition Nutrition 0.000 description 1
- 230000035764 nutrition Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000009966 trimming Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/80—Geometric correction
-
- 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
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20132—Image cropping
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30242—Counting objects in image
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The application provides a portable rice whole ear measuring method based on image recognition and application thereof, and the method comprises the following steps: s100, respectively obtaining an entire ear image of the rice and a branch image of the rice after a first branch and a second branch are separated; s200, respectively carrying out image correction on the whole ear image and the branch image to obtain a whole ear correction image and a branch correction image, and converting a scale; s300, segmenting the whole spike correction image to obtain a binary image so as to obtain the whole spike length and a path point set of the rice; s400, identifying and positioning rice spike grains; s500, acquiring and distinguishing contour information of the primary branch and the secondary branch of the rice; s600, calculating to obtain phenotype information of the primary branch and the secondary branch of the rice so as to obtain integral parameters. According to the method and the device, high-precision measurement and counting of the length, the grain number per spike and the like can be obtained, the use convenience is greatly improved, and the high-precision requirement of a user is met.
Description
Technical Field
The application relates to the technical field of plant phenotype measurement, in particular to a convenient rice whole ear measurement method based on image recognition and application thereof.
Background
Rice is one of the most important food crops in the world. The continuous increase of population puts new requirements on the breeding of crops with high yield. The rice yield is determined by three characteristics of spike number, small spike number, grain shape and the like, the small spike number is closely related to the spike branch number, and the grain size and shape influence the yield. The developmental growth of the ear of rice can also be used to assess disease state, nutrition, growth stage, etc. Therefore, the phenotypic study of whole ears of rice is necessary. However, the traditional method is time-consuming and labor-consuming for measuring the spike number, the small spike number, the grain shape and the like of the rice, and new requirements are provided for measuring the spike shape.
In the prior art, a crop seed counting method based on image processing, as disclosed in my prior patent CN201510323529.1, can accurately and quickly count seeds, and can measure parameters such as sizes and shapes of the seeds on the basis of the division, but the method needs threshing or shelling before measurement, so that the structural characteristics of ears are damaged, and information of primary and secondary branches cannot be acquired. Also for example [ D.Wu, Z.Guo, J.Ye, H.Feng, J.Liu, G.Chen, J.Zheng, D.Yan, X.Yang, X.Xiong, Q.Liu, Z.Niu, A.P.Gay, J.H.Donan, L.Xiong, W.Yang, combining high-throughput micro-CT-RGB phenotyping and genome-wide association study to segment the genetic architecture of the seed, J.Exp. Box 70 (2019) 545-561] in combination with X-ray imaging and visible light imaging to distinguish between full and unsaturated spikelets, X-ray can measure the internal structure of the seed without shelling, but the spike structure properties are still not obtained.
In summary, the existing measurement devices and measurement methods are not convenient enough in actual use, or the obtained parameters have different emphasis points. Therefore, a convenient rice whole ear measuring method based on image recognition and application thereof are urgently needed to efficiently acquire multi-parameter whole ear phenotype information.
Disclosure of Invention
The embodiment of the application provides a portable rice whole ear measuring method based on image recognition and application thereof, and aims to solve the problems that the operation in the prior art is not convenient, the overall accuracy of parameters is not good enough, and the like.
The core technology of the invention is mainly that the primary and secondary branch image acquisition of the whole ear of rice and the rice are respectively carried out on a specific background plate in sequence, the calibration object on the background plate is detected and image correction and proportional scale calculation are carried out, and the information of the length of the whole ear and the related information of the length of the branch, the grain number of the whole ear and the like are respectively obtained through the whole ear analysis and the branch analysis.
In a first aspect, the present application provides a portable method for measuring whole ears of rice based on image recognition, the method comprising the steps of:
s100, respectively obtaining an entire ear image of the rice and a branch image of the rice after separation of a primary branch and a secondary branch;
s200, respectively carrying out image correction on the whole ear image and the stem image to obtain a whole ear correction image and a stem correction image, and converting a scale;
s300, segmenting the whole panicle correction map to obtain a binary map, and performing skeleton analysis and shortest path analysis on the binary map to obtain the whole panicle length and a path point set of the rice;
s400, identifying and positioning rice spike grains based on the branch and stem correction graph;
s500, acquiring and distinguishing contour information of the primary branch and the secondary branch of the rice based on the branch correction map;
s600, calculating to obtain phenotype information of the primary branches and the secondary branches of the rice so as to obtain integral parameters and parameters of each branch.
Further, in step S100, the specific steps of the rice branch image after separating the primary branch from the secondary branch are as follows:
s110, sequentially separating all the primary branches and the secondary branches in the whole spike;
and S120, respectively placing the primary branches and the secondary branches in corresponding areas of a specific background plate, keeping the upward and downward orientation of each branch, and shooting to obtain a branch image.
Further, the specific steps of step S200 are:
s210, acquiring all possible candidate circles from the whole ear image and the branch image respectively;
s220, screening all possible candidate circles respectively, and reserving at least four candidate circles for each image;
s230, respectively calculating corresponding perspective transformation matrixes according to the reserved candidate circles;
s240, acting the perspective transformation matrix on the corresponding RGB color image to obtain a transformation image, finely adjusting the length-width ratio of the image to enable the width-height ratio of the corrected rectangular area to be consistent with the real ratio, and simultaneously recording the scale of the real size.
Further, the specific steps of step S400 are:
s410, placing different rice branches and stalks based on a specific background plate, shooting at different heights, different angles and different illuminations, collecting data containing rice spike grains, sorting and calibrating to form a database;
s420, performing offline data expansion on the database to different degrees to increase the diversity of data;
s430, performing model training based on the database and converting the model training into a specified format;
s440, dividing the branch and stem correction graph into a plurality of sub-images with overlapping, and acquiring the rectangular frame positions of all the ear grains by forward reasoning, wherein the rectangular frame positions comprise a center coordinate, a length and a width;
s450, mapping the rectangular frame position of each sub-image to the branch correction map, and removing redundant detection rectangular frames through post-processing to obtain all prediction frames.
Further, the specific steps of step S500 are:
s510, carrying out image segmentation on the branch correction graph to obtain a branch segmentation graph;
s520, obtaining all contours of the branch and stem segmentation graph by using a contour searching method, and further filtering by using a set rule to obtain an effective contour set;
s530, classifying the contour sets according to positions to respectively obtain contour sets of the primary branches and the secondary branches;
and S540, sequencing the primary branch contour set and the secondary branch contour set respectively.
Further, the specific steps of step S600 are:
s610, performing skeleton analysis and shortest path analysis on each branch to obtain a length parameter and a corresponding path point set;
s620, acquiring four vertexes of the minimum circumscribed rectangle of each branch;
s630, calculating the spike grain number of each branch and the corresponding center coordinates of all spike grains;
s640, calculating the grain attachment density of each branch, wherein the grain attachment density is the ratio of the number of grains per spike of the branch to the length of the branch;
s650, calculating phenotype information to obtain overall parameters and parameters of each branch, wherein each branch parameter comprises the length of each branch and a path point set, the minimum circumscribed rectangle of each branch, the spike grain number of each branch, the positioning center coordinates of all corresponding spike grains and the spike grain density of each branch; the overall parameters include: the total grain number, the total length of the branches, the total number of the branches, the number of primary branches, the total length of primary branches, the total grain number of primary branches, the average value of the length of primary branches, the average value of the grain number of primary branches, the contribution rate of the grain number of primary branches, the number of secondary branches, the total length of secondary branches, the total grain number of secondary branches, the average value of the length of secondary branches, the average value of the grain number of secondary branches and the contribution rate of the grain number of secondary branches.
Further, in step S600, the contribution rate of the total grain number of the primary branches is the ratio of the total grain number of the primary branches to the total grain number of the secondary branches, and the contribution rate of the total grain number of the secondary branches is the ratio of the total grain number of the secondary branches to the total grain number of the secondary branches.
In a second aspect, the present application provides a device for measuring the entire ear of rice based on image recognition, comprising:
the collection module is used for respectively obtaining an entire ear image of the rice and a branch image of the rice after a first branch and a second branch are separated;
the correction module is used for respectively carrying out image correction on the whole ear image and the branch image so as to obtain a whole ear correction image and a branch correction image and converting a scale;
the whole ear analysis module is used for segmenting the whole ear correction image to obtain a binary image, and performing skeleton analysis and shortest path analysis on the binary image to obtain the whole ear length of rice and a path point set;
the branch analysis module is used for analyzing the branch correction image, calculating the local phenotype information of each branch through spike grain detection positioning, segmentation, primary branch and secondary branch contour detection screening and further calculating the overall parameters;
the output module is used for outputting the overall parameters of the rice, and the overall parameters comprise: the total grain number, the total length of the branches, the total number of the branches, the number of primary branches, the total length of primary branches, the total grain number of primary branches, the average value of the length of primary branches, the average value of the grain number of primary branches, the contribution rate of the grain number of primary branches, the number of secondary branches, the total length of secondary branches, the total grain number of secondary branches, the average value of the length of secondary branches, the average value of the grain number of secondary branches and the contribution rate of the grain number of secondary branches.
In a third aspect, the present application provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the computer program to execute the portable rice ear-trimming measurement method based on image recognition.
In a fourth aspect, the present application provides a readable storage medium having stored therein a computer program comprising program code for controlling a process to perform a process comprising the portable image recognition-based whole ear measurement method for rice according to the above.
The main contributions and innovation points of the invention are as follows: 1. compared with the prior art, the method has the advantages that high-precision measurement and counting of the length, the grain number of the spikes and the like can be achieved only by adopting a simple background plate and a small amount of manual assistance, such as placement of the whole spikes, separation and placement of the primary branches and the secondary branches, so that the use convenience is greatly improved, and the high-precision requirement of a user is met, so that the problems of poor convenience and poor precision in the prior art are effectively solved;
2. compared with the prior art, the branch analysis in the application actually realizes the topological structure analysis of the whole ear, the number of the first-stage branches, the length and the grain attachment density of each first-stage branch, the spike grain number contribution rate of the first-stage branches, the number of the second-stage branches, the length and the grain attachment density of each second-stage branch, the spike grain number contribution rate of the second-stage branches, and the information describes the structure information of the whole ear, the attachment relation between each branch and the small spike number, the related information among all stages of branches and the like.
The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below to provide a more thorough understanding of the application.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
fig. 1 is a flow chart of a portable rice whole ear measuring method based on image recognition according to an embodiment of the application;
FIG. 2 is a schematic view showing the effect of measuring rice branches according to an embodiment of the present application;
FIG. 3 is a schematic diagram illustrating the effect of measuring the whole ears of rice according to the embodiment of the present application;
fig. 4 is a schematic diagram of a hardware structure of an electronic device according to an embodiment of the present application.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with one or more embodiments of the specification. Rather, they are merely examples of apparatus and methods consistent with certain aspects of one or more embodiments of the specification, as detailed in the claims that follow.
It should be noted that: in other embodiments, the steps of the corresponding methods are not necessarily performed in the order shown and described in this specification. In some other embodiments, the method may include more or fewer steps than those described herein. Moreover, a single step described in this specification may be broken down into multiple steps for description in other embodiments; multiple steps described in this specification may be combined into a single step in other embodiments.
The method aims at the problems that the existing measuring equipment is not convenient enough in actual use, or the obtained parameters are different in emphasis points, or the precision is low and the like, and the efficient and accurate measuring requirements of the whole ear phenotype information cannot be met.
Based on the above, the invention combines convenient equipment and advanced image recognition technology to solve the problems in the prior art.
Example one
The application aims to provide a portable rice whole ear measuring method based on an image technology, which comprises the steps of sequentially and respectively acquiring images of a whole ear of rice and primary and secondary branch stems of rice on a specific background plate, and respectively acquiring information of the length of the whole ear, the length of the branch stems, the number of the ear stems and other related information by means of two modules of whole ear analysis and branch stem analysis of a core through identifying a calibration object on the background plate, correcting the images and calculating a scale. Through manual decomposition and simple and orderly arrangement of a small number of rice branches, the whole ear image and the branch image information are directly and respectively processed and analyzed, the automation of rice whole ear measurement is realized, and the efficiency and the accuracy of whole ear phenotype measurement are improved.
Specifically, the embodiment of the present application provides a portable method for measuring the whole ears of rice based on image recognition, which may, specifically, refer to fig. 1, and comprises the following steps:
s100, respectively obtaining an entire ear image of the rice and a branch image of the rice after a first branch and a second branch are separated;
the method comprises the following specific steps of separating a primary branch image from a secondary branch image of the rice:
s110, sequentially separating all the primary branches and the secondary branches in the whole spike;
s120, respectively placing the primary branches and the secondary branches in corresponding areas of a specific background plate, keeping the upward and downward orientation of each branch, and shooting to obtain a branch image (the placing mode can be seen in figure 2);
in this embodiment, a specific background plate is selected, and then the whole ears of rice are placed on the specific background plate as required for shooting, so that a whole ear image can be obtained.
In this embodiment, the whole ear is first separated from all the primary branches by a tool, then the secondary branches on the primary branches are separated from the primary branches, and the primary branches and the secondary branches are respectively placed in corresponding regions, and each branch faces up and down, so that the starting point and the ending point of the branch and the corresponding stalk path of the starting point and the ending point on the branch are conveniently and automatically positioned in the following.
S200, respectively carrying out image correction on the whole ear image and the branch image to obtain a whole ear correction image and a branch correction image, and converting a scale;
the method comprises the following specific steps:
s210, acquiring all possible candidate circles from the whole ear image and the branch image respectively;
in this embodiment, an RGB color image is first converted into a grayscale, a local threshold segmentation method is used to obtain a binary image on the grayscale, a foreground where a pixel value is greater than 0 and a background where the value is 0 are used, and then all connected domains in the foreground are obtained by using a contour searching method, and feature calculation is performed on each connected domain, such as an area a and a minimum circumscribed circle radius r;
calculating the coincidence degree of the connected domain and the corresponding minimum circumscribed circle so as to measure the roundness of the connected domain, wherein the following formula is adopted in the embodiment:
ratio=A/(r*π 2 )
and when the ratio is greater than a given threshold value 1 and the radius r is greater than a given threshold value 2, the connected domain is considered as a candidate circle, and all possible candidate circles are obtained.
S220, screening all possible candidate circles respectively, and reserving at least four candidate circles for each image;
in this embodiment, the maximum radius maxRadius of the candidate circles is recorded, candidate circles with a radius smaller than maxRadius × 0.5 are eliminated, and four candidate circles are retained.
S230, respectively calculating corresponding perspective transformation matrixes according to the reserved candidate circles;
in the present embodiment, the four candidate circles are sorted in the order of upper left, upper right, lower right, and lower left
Simultaneously recording four same-order vertexes of a minimum right circumscribed rectangle surrounding four circle center coordinates
Establishing
To
The transmission transformation matrix warpMatrix.
S240, acting the perspective transformation matrix on the corresponding RGB color image to obtain a transformation image, finely adjusting the length-width ratio of the image to enable the width-height ratio of the corrected rectangular area to be consistent with the real ratio, and simultaneously recording the scale of the real size;
in this embodiment, the corresponding perspective transformation matrices are respectively calculated for the entire ear image (fig. 3) and the branch image, and then the perspective transformation matrices are applied to the corresponding RGB color image to obtain a transformation map, and then the length-width ratio of the image is finely adjusted, so that the width-height ratio of a rectangular region surrounded by four corrected circle centers is consistent with the true ratio and is respectively recorded as wholecorprectimg and branchcorerectimg.
Finally, the scales of pixels to the real size are recorded, namely wholeScale and branchScale respectively, which are used for converting the subsequent parameters, and the minimum circumscribed rectangles wholeRoiRect and branchRoiRect of the four calibration circles on the correlogram are recorded.
S300, segmenting the whole panicle correction image to obtain a binary image, and carrying out skeleton analysis and shortest path analysis on the binary image to obtain a whole panicle length and a path point set of rice, wherein the path point set is shown in a figure 3;
in this embodiment, the method can be based on my prior patent [202010420982, a method for measuring the length of multiple rice ears based on image processing ], where the ear length of the whole ear is defined as the length from the neck node of the ear to the top tip of the ear;
wherein, the placing requirement is that the rice ear neck section is flush with a mark line preset on the background plate. Cutting out a binary image above a mark line on the basis of cutting the length of the whole spike, carrying out pretreatment such as morphological expansion corrosion, cavity filling and the like, then obtaining a skeleton image with a foreground width of a single pixel by using a skeleton algorithm, traversing by using a graph theory shortest path algorithm to obtain paths of a starting pixel point and an ending pixel point, and selecting the starting pixel point and the ending pixel point as the lowest-end and the highest-end pixel points in the vertical direction;
preferably, a new strategy is adopted in the embodiment to select the starting point and the ending point, so that more accurate measurement is ensured. The specific operation is as follows: extracting all endpoint sets from skeleton map
Searching and counting the number of non-zero pixels in eight-connection neighborhoods around each pixel point, wherein the pixel point with the count of 1 is an end point; the starting point is determined as the point with the maximum concentration of the end points in the y direction, and is generally the position of the neck of the ear; the rice ears are sometimes bent, and the end points are concentrated into the minimum point in the y direction or the uppermost pointAnd one is real, therefore, the minimum y-direction of Nt end points is selected as a candidate end point set, a shortest Path method is utilized to obtain a Path point set Path from a starting point to each candidate end point, the longest Path is selected as the basis for calculating the final ear length, and the corresponding end point is the selected end point.
Preferably, the Path point set Path is a continuous point set along the skeleton, and since the ear-shaped contour is influenced by the placing and stretching degree, the Path has Xu Maoci, and thus, for the prior patent [202010420982, the method for measuring the lengths of multiple rice ears based on image processing]The calculation of the spike length is improved, that is, the path smoothing operation is added, which may be contour gaussian smoothing or polygonal approximation, and the embodiment is not specifically limited, and the former is adopted here; finally, sampling the smoothed path point set at certain intervals to obtain a new point set
And converting the sum of the Euclidean distances of two adjacent points by a scale, namely the measured whole ear length.
S400, identifying and positioning rice spike grains based on the branch and stem correction graph;
preferably, in order to count the total number of whole-spike rice grains and the number of corresponding whole-spike grains on each branch, an algorithm needs to be designed to identify each whole-spike grain of rice in the image. The recognition method may select a conventional machine learning technique or a deep learning technique, the latter being employed here. The deep learning technology is developed rapidly in the aspects of target detection and identification, natural language processing, big data and the like, is especially applied to the field of target detection, and provides a powerful measuring tool for agricultural breeding by means of the technologies. In this embodiment, a YOLOV5 detection framework is selected, and a mobile terminal deploys and selects TNN, which is not specifically limited in this patent.
The method mainly comprises the steps of establishing and calibrating a database, enhancing data, training a model, converting the model, carrying out a forward reasoning engine, carrying out post-processing and the like, and comprises the following specific steps:
s410, placing different rice branches and stalks based on a specific background plate, shooting at different heights, different angles and different illuminations, collecting data containing rice spike grains, sorting and calibrating to form a database;
in this embodiment, the database is established and calibrated: aiming at the specific background plate, different rice branches and stalks are placed, shooting is carried out at different heights, different angles and different illumination, meanwhile, because the number of small rice spikes is small, 2x2 cutting is carried out on each RGB image to obtain four sub images, data containing rice spike grains are collected according to the four sub images, then, third-party software LabelImg is adopted to calibrate each spike grain in each image, and then a rectangular frame and a corresponding label are established.
S420, performing offline data expansion on the database to different degrees to increase the diversity of data;
in this embodiment, data enhancement: in order to expand the coverage of data, the database in step S410 is subjected to offline data expansion of different degrees, such as rotation, illumination transformation, channel transformation, and the like, so as to increase the diversity of data and improve the generalization capability of the subsequent model trained on the basis of data.
S430, training a model based on a database and converting the model into a specified format;
in this embodiment, the model training: and (3) setting a learning rate and iteration steps by adopting YOLOV5, and starting on-line data enhancement (up-down overturning, left-right overturning, rotating, translating, zooming, mosaic and the like) to obtain a high-precision high-recall-rate model. The recall rate describes the recall rate of the model for target detection, and reflects the condition of missed detection; wherein the accuracy reflects the false detection condition of the model.
The suffix of YOLOV5 is the pth model, first converted to the intermediate model onnx, and then the TNN inference engine converted to its own specified format tnnproto/. Tnnmol.
S440, dividing the branch and stem correction graph into a plurality of sub-images with overlapping, and acquiring the rectangular frame positions of all the ear grains by forward reasoning, wherein the rectangular frame positions comprise a center coordinate, a length and a width;
in this embodiment, forward inference: and (4) carrying out forward reasoning detection by utilizing a TNN forward reasoning engine to obtain the positions of the rectangular frames of all the spike grains, including the center position, the length and the width. In particular toIn order to improve the detection accuracy of the panicle grain, the roiRect area, which is the whole background plate area in the correction map, is divided into a plurality of sub-images with overlapping, so that adjacent image blocks have repetition, for example, the overlapping rate in the horizontal and vertical directions is set. Then, each sub-image is taken as input, a TNN forward reasoning engine is used for obtaining a detection frame of each ear grain, the detection frame is restored to the coordinate on correctImg2s, and finally a comprehensive detection frame is obtained
。
S450, mapping the rectangular frame position of each sub-image to a branch correction map, and removing redundant detection rectangular frames through post-processing to obtain all prediction frames;
in the present embodiment, the post-processing: because of the division and overlapping thought of the detection area, more than one detection frame can exist in rice panicles in the overlapping area, a non-maximum suppression strategy (NMS) is adopted to remove redundant repeated detection rectangular frames for all the detection frames in the overlapping area, a high-confidence prediction frame is reserved for each target, parameters required to be set by the NMS include a confidence threshold (score) and an overlap ratio (IOU) threshold, the confidence describes the probability that the detection target is a certain type of target, the confidence threshold is used for filtering the prediction frames with the probability smaller than the value, the overlap ratio threshold is used for filtering the prediction frames with high overlap ratio, each panicle is subjected to NMS post-processing, a high-confidence prediction result is reserved, and all the last prediction frames are marked as
；
S500, as shown in FIG. 2 (limited by patent texts, the specific color in the step S510 cannot be displayed), acquiring and distinguishing contour information of the primary and secondary rice branches based on the branch correction map;
the method comprises the following specific steps:
s510, carrying out image segmentation on the branch correction graph to obtain a branch segmentation graph;
in this example, note that the image in the rectangular area branchRoiRect on branchcoredlmg is roiMat. In the above description, the corresponding areas of the first branch and the second branch are divided into two areas with different sizes, and the left and right areas are distinguished by the mark band with a certain width and a special color, the color of the mark band is selected to be blue, and in addition, since the rice ear is basically light yellow, green, and the like, in order to protrude the rice ear, the background is selected to be non-reflective black flannelette. In order to effectively separate foreground information and background such as rice ears, branches and stalks and the like, a red channel image rMat is selected as an image to be segmented in image processing. And (3) performing operations such as local adaptive threshold segmentation, morphological expansion corrosion, cavity filling and the like on the rMat to obtain a segmentation map tBon.
S520, obtaining all contours of the branch and stem segmentation graph by using a contour searching method, and further filtering by using a set rule to obtain an effective contour set;
in this embodiment, a contour search method is used for the partition map tBin to obtain all connected domains, and the circumscribed rectangle and the minimum circumscribed rectangle corresponding to each connected domain are calculated. And screening the connected domains according to a certain rule, and eliminating interference caused by factors such as impurities. The adopted rules are divided into two types, the first type of rules comprises that the length and the width of an external matrix corresponding to a connected domain are smaller than a set threshold value thrsh1, the length or the width of an external rectangle is larger than a set threshold value thrsh2, the longest side of a minimum external rectangle is smaller than a set threshold value thrsh3, the connected domains meeting any one of the rules are removed, the condition that the excessively large or small connected domains are not considered is ensured, the abnormal influence is reduced, and the remaining connected domains are collected into a set
(ii) a The second type of rule is to filter the spike count-free profile, specifically,
each connected domain of (1) i Go through
If there is one box j Is centered on the contours i If so, the connected domain is reserved, and at this time, the reserved connected domain set is recorded as
。
S530, classifying the contour sets according to positions to respectively obtain contour sets of the primary branches and the secondary branches;
in this embodiment, first, a blue mark band in the RGB image roiMat in this embodiment is detected, and a difference bMat-rMat between the blue channel bMat and the red channel rMat of the RGB image roiMat can highlight the mark band and is marked as blue featatmat, and a new segmentation map blueBin is obtained by using a fixed threshold segmentation method. The threshold is not limited here, and the embodiment is selected to be 30. And (3) acquiring all connected domains of the blueBin by adopting a contour search method, screening out the maximum connected domain according to the area, and calculating the centroid, wherein the vertical direction of the centroid is a boundary line, and the roiRect region is divided into a first region and a second region, so that when the rice branches are placed, the primary branches are positioned in the first region, and the secondary branches are positioned in the second region. Then, to
The middle arbitrary connected domain realizes the distinguishing of the outlines of the primary branch and the secondary branch according to the left-right relation between the center of the corresponding external rectangle and the boundary,
s540, sequencing the primary branch contour set and the secondary branch contour set respectively;
s600, calculating to obtain phenotype information of the primary and secondary rice branches to obtain overall parameters and parameters of each branch, wherein each branch parameter comprises the length and path point set of each branch, the minimum external rectangle of each branch, the number of spike grains of each branch, the corresponding positioning center coordinates of all the spike grains and the grain density of each branch; the overall parameters include: the total grain number, the total length of the branches, the total number of the branches, the number of primary branches, the total length of primary branches, the total grain number of primary branches, the average value of the length of primary branches, the average value of the grain number of primary branches, the contribution rate of the grain number of primary branches, the number of secondary branches, the total length of secondary branches, the total grain number of secondary branches, the average value of the length of secondary branches, the average value of the grain number of secondary branches and the contribution rate of the grain number of secondary branches.
Wherein, the contribution rate of the total grain number of the primary branches is the proportion of the total grain number of the primary branches in the total grain number of the branches, and the contribution rate of the total grain number of the secondary branches is the proportion of the total grain number of the secondary branches in the total grain number of the branches.
The method comprises the following specific steps:
s610, performing skeleton analysis and shortest path analysis on each branch to obtain a length parameter and a corresponding path point set;
in this embodiment, the idea of calculating the length parameter of the branch is similar to the obtaining of the length parameter of the whole ear and the set of path points in S300, except that the definition of the length of the branch is slightly different from the initial point in the definition of the length of the whole ear, the former is one end of the whole branch which is manually cut off, and the latter is the position of the neck node of the ear, and then the same skeleton analysis idea is adopted to determine the end point, determine the shortest path branchPath, and further perform smoothing processing, and calculate and obtain the length parameter of the branch.
S620, acquiring four vertexes of the minimum circumscribed rectangle of each branch;
in this embodiment, the minimum circumscribed rectangle of each stem contour is calculated, and then the corresponding four vertex coordinates are extracted.
S630, calculating the spike grain number of each branch and the corresponding center coordinates of all the spike grains;
frame for predicting statistical spike grain
The number of centers falling within the contour of the branch, and the coordinates of the centers of all prediction boxes belonging to the branch are recorded.
S640, calculating the grain attachment density of each branch, wherein the grain attachment density is the ratio of the number of grains per spike of the branch to the length of the branch;
s650, calculating phenotype information to obtain overall parameters and parameters of each branch.
Example two
Based on the same conception, the application also provides a whole ear of rice measuring device based on image recognition, which comprises:
the collection module is used for respectively obtaining an entire ear image of the rice and a branch image of the rice after a first branch and a second branch are separated;
the correction module is used for respectively carrying out image correction on the whole ear image and the branch image to obtain a whole ear correction image and a branch correction image and converting a scale;
the whole ear analysis module is used for segmenting the whole ear correction image to obtain a binary image, and performing skeleton analysis and shortest path analysis on the binary image to obtain the whole ear length of rice and a path point set;
the branch analysis module is used for analyzing the branch correction image, calculating the local phenotype information of each branch through spike grain detection positioning, segmentation, primary branch and secondary branch contour detection screening and further calculating the overall parameters;
the output module is used for outputting the overall parameters of the rice, and the overall parameters comprise: the total grain number, the total length of the branches, the total number of the branches, the number of primary branches, the total length of primary branches, the total grain number of primary branches, the average value of the length of primary branches, the average value of the grain number of primary branches, the contribution rate of the grain number of primary branches, the number of secondary branches, the total length of secondary branches, the total grain number of secondary branches, the average value of the length of secondary branches, the average value of the grain number of secondary branches and the contribution rate of the grain number of secondary branches.
EXAMPLE III
The present embodiment also provides an electronic device, referring to fig. 4, comprising a memory 404 and a processor 402, wherein the memory 404 stores a computer program, and the processor 402 is configured to execute the computer program to perform the steps of any of the above method embodiments.
Specifically, the processor 402 may include a Central Processing Unit (CPU), or A Specific Integrated Circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.
The processor 402 reads and executes the computer program instructions stored in the memory 404 to implement any convenient method for measuring the whole rice ears based on image recognition in the above embodiments.
Optionally, the electronic apparatus may further include a transmission device 406 and an input/output device 408, where the transmission device 406 is connected to the processor 402, and the input/output device 408 is connected to the processor 402.
The transmitting device 406 may be used to receive or transmit data via a network. Specific examples of the network described above may include wired or wireless networks provided by communication providers of the electronic devices. In one example, the transmission device includes a Network adapter (NIC) that can be connected to other Network devices through a base station to communicate with the internet. In one example, the transmitting device 406 may be a Radio Frequency (RF) module, which is used to communicate with the internet in a wireless manner.
The input and output devices 408 are used to input or output information. In this embodiment, the input information may be etc., and the output information may be etc.
Example four
The embodiment also provides a readable storage medium, in which a computer program is stored, the computer program including program code for controlling a process to execute the process, the process including the portable image recognition-based whole rice ear measurement method according to the first embodiment.
It should be noted that, for specific examples in this embodiment, reference may be made to examples described in the foregoing embodiments and optional implementations, and details of this embodiment are not described herein again.
In general, the various embodiments may be implemented in hardware or special purpose circuits, software, logic or any combination thereof. Some aspects of the invention may be implemented in hardware, while other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device, although the invention is not limited thereto. While various aspects of the invention may be illustrated and described as block diagrams, flow charts, or using some other pictorial representation, it is well understood that these blocks, apparatus, systems, techniques or methods described herein may be implemented in, as non-limiting examples, hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.
Embodiments of the invention may be implemented by computer software executable by a data processor of the mobile device, such as in a processor entity, or by hardware, or by a combination of software and hardware. Computer software or programs (also referred to as program products) including software routines, applets and/or macros can be stored in any device-readable data storage medium and they include program instructions for performing particular tasks. The computer program product may comprise one or more computer-executable components configured to perform embodiments when the program is run. The one or more computer-executable components may be at least one software code or a portion thereof. Further in this regard it should be noted that any block of the logic flow as in the figures may represent a program step, or an interconnected logic circuit, block and function, or a combination of a program step and a logic circuit, block and function. The software may be stored on physical media such as memory chips or memory blocks implemented within the processor, magnetic media such as hard or floppy disks, and optical media such as, for example, DVDs and data variants thereof, CDs. The physical medium is a non-transitory medium.
It should be understood by those skilled in the art that various features of the above embodiments can be combined arbitrarily, and for the sake of brevity, all possible combinations of the features in the above embodiments are not described, but should be considered as within the scope of the present disclosure as long as there is no contradiction between the combinations of the features.
The above examples only express several embodiments of the present application, and the description thereof is more specific and detailed, but not to be construed as limiting the scope of the present application. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.
Claims (10)
1. The convenient and fast rice whole ear measuring method based on image recognition is characterized by comprising the following steps of:
s100, respectively obtaining an entire ear image of the rice and a branch image of the rice after a first branch and a second branch are separated;
s200, respectively carrying out image correction on the whole ear image and the branch image to obtain a whole ear correction image and a branch correction image, and converting a scale;
s300, segmenting the whole ear correction map to obtain a binary map, and performing skeleton analysis and shortest path analysis on the binary map to obtain the whole ear length of rice and a path point set;
s400, identifying and positioning rice spike grains based on the branch and stalk correction map;
s500, acquiring and distinguishing contour information of the primary branch and the secondary branch of the rice based on the branch correction map;
s600, calculating to obtain phenotype information of the primary branch and the secondary branch of the rice so as to obtain an integral parameter and each branch parameter.
2. The portable rice whole ear measuring method based on image recognition as claimed in claim 1, wherein in step S100, the specific steps of the branch image after the rice is separated into the first branch and the second branch are as follows:
s110, sequentially separating all the primary branches and the secondary branches in the whole spike;
and S120, respectively placing the primary branches and the secondary branches in corresponding areas of a specific background plate, keeping the upward and downward orientation of each branch, and shooting to obtain a branch image.
3. The portable rice whole ear measuring method based on image recognition according to claim 1, wherein the step S200 comprises the following steps:
s210, acquiring all possible candidate circles from the whole ear image and the branch image respectively;
s220, screening all possible candidate circles respectively, and reserving at least four candidate circles for each image;
s230, respectively calculating corresponding perspective transformation matrixes according to the reserved candidate circles;
s240, acting the perspective transformation matrix on the corresponding RGB color image to obtain a transformation image, finely adjusting the length-width ratio of the image to enable the width-height ratio of the corrected rectangular area to be consistent with the real ratio, and simultaneously recording the scale of the real size.
4. The portable rice whole ear measuring method based on image recognition according to claim 1, wherein the step S400 comprises the following steps:
s410, placing different rice branches and stalks based on a specific background plate, shooting at different heights, different angles and different illuminations, collecting data containing rice spike grains, sorting and calibrating to form a database;
s420, performing offline data expansion on the database to different degrees to increase the diversity of data;
s430, performing model training based on the database and converting the model training into a specified format;
s440, dividing the branch and stalk correction graph into a plurality of sub-images with overlapping, and acquiring the rectangular frame positions of all the ear grains by forward reasoning, wherein the rectangular frame positions comprise a center coordinate, a length and a width;
s450, mapping the rectangular frame position of each sub-image to the branch correction map, and removing redundant detection rectangular frames through post-processing to obtain all prediction frames.
5. The portable rice whole ear measuring method based on image recognition according to claim 1, wherein the step S500 comprises the following steps:
s510, carrying out image segmentation on the branch correction graph to obtain a branch segmentation graph;
s520, obtaining all contours of the branch and stem segmentation graph by using a contour searching method, and further filtering by using a set rule to obtain an effective contour set;
s530, classifying the contour sets according to positions to respectively obtain contour sets of the primary branches and the secondary branches;
and S540, sequencing the primary branch contour set and the secondary branch contour set respectively.
6. The portable rice whole ear measuring method based on image recognition according to claim 5, wherein the step S600 comprises the following steps:
s610, performing skeleton analysis and shortest path analysis on each branch to obtain a length parameter and a corresponding path point set;
s620, acquiring four vertexes of the minimum circumscribed rectangle of each branch;
s630, calculating the spike grain number of each branch and the corresponding center coordinates of all the spike grains;
s640, calculating the heading density of each branch, wherein the heading density is the ratio of the number of grains of the branch to the length of the branch;
s650, calculating phenotype information to obtain overall parameters and parameters of each stem, wherein each stem parameter comprises the length of each stem and a path point set, the minimum external rectangle of each stem, the spike grain number of each stem, the positioning center coordinates of all corresponding spike grains and the spike grain density of each stem; the overall parameters include: the total grain number, the total length of the branches, the total number of the branches, the number of primary branches, the total length of primary branches, the total grain number of primary branches, the average value of the length of primary branches, the average value of the grain number of primary branches, the contribution rate of the grain number of primary branches, the number of secondary branches, the total length of secondary branches, the total grain number of secondary branches, the average value of the length of secondary branches, the average value of the grain number of secondary branches and the contribution rate of the grain number of secondary branches.
7. The portable image-recognition-based rice whole-ear measurement method according to claim 1, wherein in step S600, the contribution rate of the total grain number of the primary stalks is the ratio of the total grain number of the primary stalks to the total grain number of the stalks, and the contribution rate of the total grain number of the secondary stalks is the ratio of the total grain number of the secondary stalks to the total grain number of the stalks.
8. A whole ear of grain measuring device of rice based on image recognition, its characterized in that includes:
the collection module is used for respectively obtaining an entire ear image of the rice and a branch image of the rice after a first branch and a second branch are separated;
the correction module is used for respectively carrying out image correction on the whole ear image and the branch image to obtain a whole ear correction image and a branch correction image and converting a scale;
the whole ear analysis module is used for segmenting the whole ear correction map to obtain a binary map, and performing skeleton analysis and shortest path analysis on the binary map to obtain the whole ear length of rice and a path point set;
the branch analysis module is used for analyzing the branch correction image, calculating the local phenotype information of each branch through spike grain detection positioning, segmentation, primary branch and secondary branch contour detection screening and further calculating the overall parameters;
the output module is used for outputting the overall parameters of the rice, and the overall parameters comprise: the total grain number, the total length of the branches, the total number of the branches, the number of primary branches, the total length of primary branches, the total grain number of primary branches, the average value of the length of primary branches, the average value of the grain number of primary branches, the contribution rate of the grain number of primary branches, the number of secondary branches, the total length of secondary branches, the total grain number of secondary branches, the average value of the length of secondary branches, the average value of the grain number of secondary branches and the contribution rate of the grain number of secondary branches.
9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the computer program to perform the portable image recognition-based rice panicle measurement method according to any one of claims 1 to 7.
10. A readable storage medium having stored therein a computer program comprising program code for controlling a process to execute a process comprising the portable image recognition-based rice panicle measurement method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211322966.8A CN115375694A (en) | 2022-10-27 | 2022-10-27 | Portable rice whole ear measuring method based on image recognition and application thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211322966.8A CN115375694A (en) | 2022-10-27 | 2022-10-27 | Portable rice whole ear measuring method based on image recognition and application thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115375694A true CN115375694A (en) | 2022-11-22 |
Family
ID=84072840
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211322966.8A Pending CN115375694A (en) | 2022-10-27 | 2022-10-27 | Portable rice whole ear measuring method based on image recognition and application thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115375694A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116423054A (en) * | 2023-03-09 | 2023-07-14 | 中铁九桥工程有限公司 | U rib plate welding method and welding system |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010088407A (en) * | 2008-10-03 | 2010-04-22 | Honda Motor Co Ltd | Method and apparatus for analyzing structure of ear |
| CN103632157A (en) * | 2012-08-24 | 2014-03-12 | 南京农业大学 | A method for counting seeds of a wheatear portion per wheat |
| CN104021369A (en) * | 2014-04-30 | 2014-09-03 | 南京农业大学 | Grain counting method for spike of single rice based on digital image processing technology |
| CN105115469A (en) * | 2015-07-29 | 2015-12-02 | 华中农业大学 | Paddy rice spike phenotypic parameter automatic measuring and spike weight predicting method |
| CN106971393A (en) * | 2017-02-23 | 2017-07-21 | 北京农业信息技术研究中心 | The phenotype measuring method and system of a kind of corn kernel |
| CN109557090A (en) * | 2017-09-23 | 2019-04-02 | 华中农业大学 | The lossless extraction rice spike of rice character of X-ray-visible light Double-mode imaging |
| CN109738442A (en) * | 2019-01-05 | 2019-05-10 | 华中农业大学 | A kind of full-automatic extraction system of rice spike of rice character based on the registration imaging of big view X-ray visible light |
| CN111724354A (en) * | 2020-06-02 | 2020-09-29 | 浙江托普云农科技股份有限公司 | Image processing-based method for measuring spike length and small spike number of multiple wheat |
| CN111738936A (en) * | 2020-05-18 | 2020-10-02 | 浙江托普云农科技股份有限公司 | Image processing-based multi-plant rice spike length measuring method |
| CN112164030A (en) * | 2020-09-04 | 2021-01-01 | 华南农业大学 | Method and device for quickly detecting rice panicle grains, computer equipment and storage medium |
| CN113012150A (en) * | 2021-04-14 | 2021-06-22 | 南京农业大学 | Feature-fused high-density rice field unmanned aerial vehicle image rice ear counting method |
| CN113470104A (en) * | 2021-07-21 | 2021-10-01 | 浙江托普云农科技股份有限公司 | Wheat ear grain counting and measuring method and system based on gray value comparison |
| CN113628186A (en) * | 2021-08-09 | 2021-11-09 | 海南青峰生物科技有限公司 | System for measuring rice spike number based on 5G communication and image recognition |
| CN114066842A (en) * | 2021-11-12 | 2022-02-18 | 浙江托普云农科技股份有限公司 | Method, system and device for counting number of ears and storage medium |
-
2022
- 2022-10-27 CN CN202211322966.8A patent/CN115375694A/en active Pending
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010088407A (en) * | 2008-10-03 | 2010-04-22 | Honda Motor Co Ltd | Method and apparatus for analyzing structure of ear |
| CN103632157A (en) * | 2012-08-24 | 2014-03-12 | 南京农业大学 | A method for counting seeds of a wheatear portion per wheat |
| CN104021369A (en) * | 2014-04-30 | 2014-09-03 | 南京农业大学 | Grain counting method for spike of single rice based on digital image processing technology |
| CN105115469A (en) * | 2015-07-29 | 2015-12-02 | 华中农业大学 | Paddy rice spike phenotypic parameter automatic measuring and spike weight predicting method |
| CN106971393A (en) * | 2017-02-23 | 2017-07-21 | 北京农业信息技术研究中心 | The phenotype measuring method and system of a kind of corn kernel |
| CN109557090A (en) * | 2017-09-23 | 2019-04-02 | 华中农业大学 | The lossless extraction rice spike of rice character of X-ray-visible light Double-mode imaging |
| CN109738442A (en) * | 2019-01-05 | 2019-05-10 | 华中农业大学 | A kind of full-automatic extraction system of rice spike of rice character based on the registration imaging of big view X-ray visible light |
| CN111738936A (en) * | 2020-05-18 | 2020-10-02 | 浙江托普云农科技股份有限公司 | Image processing-based multi-plant rice spike length measuring method |
| CN111724354A (en) * | 2020-06-02 | 2020-09-29 | 浙江托普云农科技股份有限公司 | Image processing-based method for measuring spike length and small spike number of multiple wheat |
| CN112164030A (en) * | 2020-09-04 | 2021-01-01 | 华南农业大学 | Method and device for quickly detecting rice panicle grains, computer equipment and storage medium |
| CN113012150A (en) * | 2021-04-14 | 2021-06-22 | 南京农业大学 | Feature-fused high-density rice field unmanned aerial vehicle image rice ear counting method |
| CN113470104A (en) * | 2021-07-21 | 2021-10-01 | 浙江托普云农科技股份有限公司 | Wheat ear grain counting and measuring method and system based on gray value comparison |
| CN113628186A (en) * | 2021-08-09 | 2021-11-09 | 海南青峰生物科技有限公司 | System for measuring rice spike number based on 5G communication and image recognition |
| CN114066842A (en) * | 2021-11-12 | 2022-02-18 | 浙江托普云农科技股份有限公司 | Method, system and device for counting number of ears and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| M. V. C. PADILLA ET AL: "Rice Spikelet Yield Determination Using Image Processing", 《2018HNICEM》 * |
| 翟雪: "基于大田水稻穗部图像特征的测产技术研究", 《中国优秀硕士学位论文全文数据库农业科技辑》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116423054A (en) * | 2023-03-09 | 2023-07-14 | 中铁九桥工程有限公司 | U rib plate welding method and welding system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114120037B (en) | Germinated potato image recognition method based on improved yolov5 model | |
| CN114170148A (en) | Corn plant type parameter measuring method, system, device and storage medium | |
| CN112069985B (en) | High-resolution field image rice spike detection and counting method based on deep learning | |
| CN114782838B (en) | Rice identification method and device, electronic equipment and storage medium | |
| CN114066842A (en) | Method, system and device for counting number of ears and storage medium | |
| CN115512232B (en) | Crop seed germination condition identification model, construction method and application thereof | |
| CN115620151B (en) | Method and device for identifying phenological period, electronic equipment and storage medium | |
| CN113223040B (en) | Banana estimated yield method and device based on remote sensing, electronic equipment and storage medium | |
| CN115375694A (en) | Portable rice whole ear measuring method based on image recognition and application thereof | |
| Lootens et al. | High-throughput phenotyping of lateral expansion and regrowth of spaced Lolium perenne plants using on-field image analysis | |
| CN110751035A (en) | Seed corn production identification method and device | |
| CN116229265A (en) | Method for automatically and nondestructively extracting phenotype of soybean plants | |
| CN108961295B (en) | Purple soil image segmentation and extraction method based on normal distribution H threshold | |
| CN114419367A (en) | High-precision crop drawing method and system | |
| CN111738310B (en) | Material classification method, device, electronic equipment and storage medium | |
| CN111582035B (en) | Fruit tree age identification method, device, equipment and storage medium | |
| CN115953686B (en) | Peanut pest detection method and system based on image processing | |
| CN114782837B (en) | Plant estimation method, plant estimation device, electronic equipment and storage medium | |
| CN114782455B (en) | Cotton row center line image extraction method for agricultural machine embedded equipment | |
| CN114782835A (en) | Crop lodging area proportion detection method and device | |
| CN113807129A (en) | Crop area identification method and device, computer equipment and storage medium | |
| Doraiswamy et al. | Improved techniques for crop classification using MODIS imagery | |
| Zhu et al. | Stem-leaf segmentation and phenotypic trait extraction of maize shoots from three-dimensional point cloud | |
| CN116052141B (en) | Crop growth period identification method, device, equipment and medium | |
| Mudgil et al. | Identification of Tomato Plant Diseases Using CNN-A Comparative Review |
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 |