JP5180772B2 - Method and apparatus for analyzing ear structure - Google Patents

Description

  The present invention relates to a method and an apparatus for analyzing a branch structure of a plant ear.

  In the breeding of plants including crops such as rice, QTL (Quantitative) that estimates the location of genes related to the properties in the chromosome by quantifying useful properties such as plant morphology and comparing them with genotypes Trait Locus) analysis is used. For example, in the QTL analysis of rice, properties such as the shape of the ear of rice, the number of grains of rice bran and the date of heading are used as parameters.

  However, since the shape of the rice ear is complicated, conventionally, measurement for quantifying the rice ear shape as a parameter had to be performed manually. Specifically, by manually expanding a single ear, the number of branches, the branching state of the branches, the length of the branches, the number of folds, the heels and sterility of the heels were measured. . Such work is very time consuming and requires skill.

In recent years, an apparatus for measuring and sorting the shape of a ridge has been developed (for example, Patent Document 1 and Patent Document 2). However, rice buds have a three-dimensional and complex structure, unlike cocoons. As a method for analyzing a three-dimensional structure, there are methods using X-ray CT (Computed Tomography) and MRI (Magnetic Resonance Imaging). However, the apparatus becomes large and the amount of data to be processed becomes enormous, which is not general. As described above, a general-purpose analysis method and analysis apparatus for analyzing the structure of ears such as rice ears including branching states of branches have not been developed.
JP 2002-312762 A JP 2005-55245 A

  Therefore, there is a need for a general-purpose analysis method and analysis device for analyzing the morphology of a spike including a branching structure of a spike such as a rice spike.

  The analysis method according to the present invention is an analysis method for analyzing the branch structure of a plant ear, the step of collecting an image of the ear, the step of extracting branches and wrinkles from the image of the ear, and the state of branching of the branch Defining the steps. The step of determining the branching state of the branch includes storing the branching state of the branch in a data file having a tree structure corresponding to the branching state of the branch.

  An analysis apparatus according to the present invention is an analysis apparatus that analyzes a branch structure of a plant ear, an image collection unit that collects an image of the ear, a partial extraction unit that extracts branches and wrinkles from the image of the ear, A branch state determination unit for determining the branch state of the branch, and a data file having a tree structure corresponding to the branch state of the branch. The branch state determination unit stores the branch state of the branch in the data file.

  According to the present invention, since the data file having a tree structure corresponding to the branching state of the branch is used, it is possible to grasp the branching state of the ear branch without using complicated logic. Therefore, according to the present invention, there is provided a general-purpose analysis method and analysis apparatus for analyzing the morphology of ears such as rice ears including branching states of branches.

  The features of the embodiment of the present invention are as follows.

  In the analysis method according to the embodiment of the present invention, the step of determining the branching state of the branch sets the ear root at the upstream, defines an initial start point on the branch near the ear root, and downstream of the initial start point The end point of the branch adjacent to the first start point, which is the branching point of the branch or the point where the wrinkle is attached, is the end component from the first starting point to the end point, and when the end point is a branch point, the end point is Another branch component is defined as a new starting point, thus dividing the entire spike branch into branch components.

  According to the present embodiment, by dividing the entire branch of the spike into branch components, it becomes easy to analyze the structure of the spike including the branching state of the branch.

  In the analysis method according to the embodiment of the present invention, the data file includes point tags corresponding to all branch points and end points, and the point tag corresponding to the branch point corresponds to an end point of the branch component starting from the branch point. The point tag corresponding to all the branch points and end points includes length information from the adjacent upstream branch point, and forms a tree structure.

  According to the present embodiment, the point tag corresponding to the branch point and the end point is combined in a parent-child relationship, that is, a hierarchical relationship in which the parent is higher and the child is lower, so that the tree structure corresponding to the branch state of the branch is obtained. A data file can be easily created.

  In the analysis method according to the embodiment of the present invention, a point tag corresponding to a branch point or an end point has information on the branch point or the end point and information on a branch component whose end point is the branch point or the end point.

  According to the present embodiment, branch information can be easily stored in a data file having a tree structure corresponding to the branch branch state.

  In the analysis method according to the embodiment of the present invention, the point tag corresponding to the end point has information on the eyelid as a child element.

  According to the present embodiment, information on a cocoon can be easily stored in a data file having a tree structure corresponding to the branching state of the branches.

  In the analysis method according to the embodiment of the present invention, the cocoon information includes trait information representing the state of the cocoon as a child element.

  According to the present embodiment, it is possible to utilize the trait information representing the state of the cocoon in combination with information related to the structure of the ear.

  In the analysis method according to the embodiment of the present invention, the step of determining the branch state of the branch includes, as a main branch downstream from the start point, a branch composed of a branch component connecting the end point having the longest distance from the start point and the start point. , Including branching from the main branch as a branch of the main branch.

  According to the present embodiment, it is possible to automatically determine the main branch and the branch branch.

  In the analysis method according to the embodiment of the present invention, the data file includes point tags corresponding to all branch points and end points, and the point tag corresponding to the branch point corresponds to an end point of the branch component starting from the branch point. The point tag corresponding to the branch point or the end point has information on the branch point or the end point and information on a branch component having the branch point or the end point as an end point. The information includes information on whether the branch component is a main branch or a branch branch.

  According to the present embodiment, it is possible to utilize information regarding whether the branch component is a main branch or a branch branch in combination with information regarding the structure of other spikes.

  FIG. 1 is a diagram showing a configuration of an analysis apparatus according to an embodiment of the present invention. The analysis apparatus includes an image collection unit 101 that collects an image of a spike, a partial extraction unit 103 that extracts a portion including a spike branch and a cocoon, a branch state determination unit 105 that determines a branch state of the spike branch, and a post-processing A data management unit 107 that stores the data and an input / output unit 109 that manages input / output processing. The analysis apparatus further includes a work file storage unit 111, an array variable file storage unit 113, a form data file storage unit 115, and a save file storage unit 117 as storage devices. The image collection unit 101 may be configured from a scanner. The partial extraction unit 103, the branch state determination unit 105, the data management unit 107, the input / output unit 109, and each storage unit may be composed of one or more computers such as a personal computer.

  FIG. 2 is a flowchart illustrating an analysis method according to an embodiment of the present invention. Step S010 in FIG. 2 is performed by the image collection unit 101. Steps S020 to S100 in FIG. 2 are performed by the partial extraction unit 103. Step S110 in FIG. 2 is performed by the branch state determination unit 105. Step S120 in FIG. 2 is performed by the data management unit 107.

  In step S010 in FIG. 2, the image collection unit 101 reads a spike image. Prior to loading the ear image, the ear is unfolded and manually pasted onto a transparent film so that the branches and folds do not overlap. The image collecting unit 101 is composed of a transmissive scanner. The resolution of the scanner is preferably 300 to 600 dpi (dpi is the number of pixels per inch). If the resolution is made higher, the measurement resolution is improved, but the scan time and analysis time become longer and the memory load becomes larger. The reason for using a transmissive scanner is that when a reflective scanner is used, shadows are generated, causing measurement errors and reading errors. The image collection unit 101 stores the read image in the work file 111.

  In step S020 in FIG. 2, the partial extraction unit 103 binarizes the image read by the image collection unit 101. The partial extraction unit 103 stores the binarized image in the work file 111.

  FIG. 15 is a diagram illustrating an example of a binarized ear image.

  FIG. 3 is a flowchart showing the detailed processing of step S030 in FIG.

  In step S03010 of FIG. 3, the partial extraction unit 103 separates the heel part of the ear from the branch part in the binarized ear image. You may perform the process which isolate | separates the buttock part of a head from a branch part using the isolation | separation processing function contained in commercially available image processing software. For example, the buttocks may be separated from the branches by contraction / expansion processing. The contraction / expansion process is described in, for example, Japanese Patent Application Laid-Open No. 2005-83775 as a process for selecting a grain.

  In step S03020 of FIG. 3, the partial extraction unit 103 determines whether the size of the buttock is within a predetermined range. In order to determine whether the size of the buttock is within a predetermined range, for example, the following method may be used. The circumscribed circle and the inscribed circle of each buttock area (region) are obtained. Here, the circumscribed circle is the smallest circle including the entire buttock area, and the inscribed circle is the largest circle included in the buttock area. The diameter of the circumscribed circle corresponds to the length of the ridge. The diameter of the inscribed circle corresponds to the width of the ridge. As an example, it is assumed that the circumscribed circle has a diameter of 10 mm or less and the inscribed circle has a diameter of 1 mm or more within a predetermined range. If the size of the buttock is within the predetermined range, the process proceeds to step S03030. If the size of the buttock is not within the predetermined range, the process proceeds to step S03040.

  In step S03030 of FIG. 3, the partial extraction unit 103 stores the data of the buttock.

  In step S03040 of FIG. 3, the partial extraction unit 103 deletes the data of the heel from the area.

  FIG. 16 is a diagram illustrating an example of the extracted image of the buttock. The partial extraction unit 103 stores the extracted collar image in the work file 111.

  FIG. 4 is a flowchart showing the detailed process of step S040 of FIG.

  In step S04010 of FIG. 4, the partial extraction unit 103 removes the heel from the binarized ear image.

  In step S04020 of FIG. 4, the partial extraction unit 103 determines whether the branch size such as the thickness of the branch is within a predetermined range. If the branch size is within the predetermined range, the process proceeds to step S04030. If the branch size is not within the predetermined range, the process proceeds to step S04040.

  In step S04030 of FIG. 4, the partial extraction unit 103 stores the branch data.

  In step S04040 of FIG. 4, the partial extraction unit 103 erases the data of the branch from the area.

  In step S04050 of FIG. 4, the partial extraction unit 103 performs a rounding (smoothing) process on the stored branch data.

  FIG. 17 is a diagram illustrating an example of the extracted branch image. The partial extraction unit 103 stores the extracted branch image in the work file 111.

  FIG. 5 is a flowchart showing the detailed process of step S050 of FIG.

  In step S05010 of FIG. 5, the partial extraction unit 103 performs branch thinning. “Thinning” means that information on the thickness of the branch portion is ignored and represented by a predetermined line.

  In step S05020 in FIG. 5, the partial extraction unit 103 extracts the branch points and end points of the branches. Here, a branch point is a point where a branch is branched, and an end point is a point where a tip of the branch should be wrinkled.

  FIG. 6 is a flowchart showing the detailed processing of step S060 in FIG.

  In step S06010 of FIG. 6, the partial extraction unit 103 removes the branch point from the branch part and divides the branch part into each branch component. Here, the branch component is a section of a branch having two adjacent branch points or adjacent branch points and end points as both ends.

  In step S06020 in FIG. 6, the partial extraction unit 103 measures the length of each branch component based on the size of the pixel.

  In step S06030 of FIG. 6, the partial extraction unit 103 stores the number of each branch component, the coordinates and lengths of both end points in the array variable file 113.

  In step S070 in FIG. 2, the input / output unit 109 assists the operator in editing the image data stored in the work file 111. Editing operations include point deletion, point addition, line deletion, line addition, heel deletion, heel addition, branch point to end point change, and end point to branch point change. Specifically, the input / output unit 109 has a function of displaying an image stored in the work file 111 and a function of supporting an operator's editing work using the displayed image. More specifically, for example, editing processing may be performed by clicking an icon indicating each editing support function and an editing target on the screen displaying the image with a mouse. In response to an operator instruction via the input / output unit 109, the partial extraction unit 103 and the branch state determination unit 105 perform reprocessing such as recalculation and redrawing.

  In step S080 of FIG. 2, the input / output unit 109 assists the operator in specifying the head and head. Here, the head of the head is a boundary point of the branch component at the tip of the head. The head is the boundary point of the branch component at the base of the ear and is the starting point of the structural analysis of the ear, as will be described later. The region from the neck to the top is called the cob. Since the diameter of the branch is slightly thicker, the head of the head and the head of the head may be automatically specified by image processing. However, in the present embodiment, the operator can specify the head. Specifically, the input / output unit 109 displays, for example, an image extracted from the branch part stored in the work file 111, and the operator clicks the point indicating the head and the head on the image. It may be possible to specify by doing so. The input / output unit 109 stores in the array variable file 113 the position data of the head and neck specified by the operator.

  FIG. 7 is a flowchart showing the detailed process of step S090 of FIG.

  In step S09010 of FIG. 7, the partial extraction unit 103 detects wrinkles near the end points.

  In step S09020 of FIG. 7, the partial extraction unit 103 determines whether the distance between the end point and the eyelid is within a predetermined range. If the distance between the end point and the eyelid is within a predetermined range, the process proceeds to step S09030. If the distance between the end point and the heel is not within the predetermined range, the process proceeds to step S09040.

  In step S09030 of FIG. 7, the partial extraction unit 103 combines the end point and the ridge.

  In step S09050 of FIG. 7, the partial extraction unit 103 stores the result of the combination in the array variable file 113. Specifically, in the array variable file 113, the numbers and coordinates of the ridges are associated with the branch component numbers.

  In step S09040 in FIG.

  In step S09060 in FIG. 7, the partial extraction unit 103 determines whether there is an unprocessed endpoint. If there is no unprocessed end point, the process proceeds to step S09070. If there is an unprocessed end point, the process returns to step S09010.

  In step S09070 in FIG. 7, the partial extraction unit 103 obtains all pixel average values of gray scale values of the buttock area for all the buttock areas.

  In step S09080 in FIG. 7, the partial extraction unit 103 obtains a threshold value of the average value of all pixels by a discriminant analysis method. The threshold value is determined so that the variance between the two classes is maximized when the buttock is classified into two classes according to the threshold value. Regarding discriminant analysis, for example, http://www.mm.media.kyoto-u.ac.jp/education/DIP/WEBPAGE_SECTION/section7/node2.html and Japanese Patent No. 4129598 (paragraphs 38 to 423 and (Paragraph 65).

  In step S09090 of FIG. 7, if the average value of all pixels of the gray scale value in the buttock area exceeds the threshold value, the partial extraction unit 103 sets “籾” as the 籾 and the average value of all pixels of the gray scale value of the buttock area. If is less than or equal to the threshold value, the bag is determined to be “sterile”. The “稔” state of the cocoon refers to a state where the fruit is present in the rice husk, and the “non-sterile” state refers to a state where the fruit does not exist in the rice husk.

  In step S09100 in FIG. 7, the partial extraction unit 103 stores the result of “稔” or “not obscured” in the array variable file 113.

  FIG. 8 is a flowchart showing the detailed process of step S100 of FIG.

  In step S10010 of FIG. 8, the partial extraction unit 103 detects a branch point or an end point near both ends of the branch component.

  In step S10020 of FIG. 8, the partial extraction unit 103 combines the branch component and the branch point or end point.

  In step S10030 of FIG. 8, the partial extraction unit 103 stores the result of the combination in the array variable file 113. Specifically, in the array variable file 113, the branch point or end point number and coordinates are associated with the branch component number.

  In step S10040 of FIG. 8, the partial extraction unit 103 determines whether there is an unprocessed branch component. If there is no unprocessed branch component, the process ends. If there is an unprocessed branch component, the process returns to step S10010.

  9 to 11 are flowcharts showing the detailed processing of step S110 of FIG.

  In step S11010 of FIG. 9, the branch state determination unit 105 designates the file name of the configuration data file 115. The branch state determination unit 105 performs processing using the data of the array variable file 113 created by the partial extraction unit 103 and creates a form data file 115. The form data file 115 includes point tags corresponding to branch points and end points as components, and the point tags are associated with each other. Each point tag has information. The form data file 115 is preferably created as an XML file. The XML file is a tree-structured data file. The XML file is described in, for example, Japanese Patent No. 3492247.

  In step S11020 of FIG. 9, the branch state determination unit 105 starts from the head. Here, the starting point is a point at which the ear analysis process is started. The starting point is the most upstream point, and the direction away from the starting point is the downstream direction.

  In step S11030 of FIG. 9, the branching state determining unit 105 sets the starting point as the starting point A and creates a point tag for the starting point A. Here, the start point is a point at the upstream end of the branch component. The branch state determination unit 105 adds the attribute value of the start point A to the point tag of the start point A. The attribute value includes the coordinates of the starting point A. The attribute value data can be acquired from the array variable file 113.

  In step S11040 of FIG. 9, the branch state determination unit 105 determines all branch components including the start point A and located downstream from the start point A.

  In step S11050 of FIG. 9, the branch state determination unit 105 sets the end points other than the start point A of each branch component as end points B (B1, B2,... Bn). Here, n is the number of all branch components including the start point A and located downstream from the start point A.

  In step S11060 of FIG. 9, the branch state determination unit 105 determines the branch components of the start point A and the end point B (B1, B2,... Bn).

  In step S11070 of FIG. 9, the branch state determination unit 105 creates a point tag of the end point B (B1, B2,... Bn) as a child element of the point tag of the start point A in the configuration data file 115.

  In step S11080 of FIG. 9, the branch state determination unit 105 adds the attribute values of the end point B and the branch component AB to the point tag of the end point B (B1, B2,... Bn). The attribute value includes the coordinates of the end point B and the length of the branch component AB. The attribute value data can be acquired from the array variable file 113.

  In step S11090 of FIG. 9, the branch state determination unit 105 determines whether or not the end point B (B1, B2,... Bn) is an end point. If the end point B is an end point, the process proceeds to step S11110. If the end point B is not an end point, the process proceeds to step S11100.

  In step S11100 of FIG. 9, the branch state determination unit 105 sets the end point B as the next start point A and proceeds to step S11040.

  In step S11110 of FIG. 9, the branch state determination unit 105 adds the end point information to the end point B point tag.

  Through the above processing, point tags of all branch components from the starting point to the end point are created in the configuration data file 115.

  In step S11120 of FIG. 9, the branch state determination unit 105 sets the end point B as the next start point A and proceeds to step S11130.

  In step S11130 of FIG. 10, the branch state determination unit 105 searches for a bag connected to the start point A that is an end point.

  In step S11140 of FIG. 10, the branch state determination unit 105 determines whether the eyelid is connected to another branch component. If the kite is connected to another branch component, the process proceeds to step S11160. If it does not lead to another branch component, the process proceeds to step S11150.

  In step S11160 of FIG. 10, the branch state determination unit 105 creates a branch tag as a child element of the point tag of the start point A. Here, the branch tag is a tag for designating the file name of another file representing the branch portion. If there is a lot of branching of the ears and the branching structure of the ears cannot be expanded on a single sheet, the primary branch is cut off from the cob and analyzed separately (measured), and the analysis results are stored in another file (morphological data file). To do. Use a branch tag for that.

  In step S11150 of FIG. 10, the branch state determination unit 105 creates a grain tag as a child element of the point tag of the start point A. A grain tag is a child element of a point tag at an end point, and is a tag having information on a bag.

  In step S11170 of FIG. 10, the branch state determination unit 105 determines whether there is a defect. If there is a defect, the process proceeds to step S11190. If there is no defect, the process proceeds to step S11180.

  In step S11180 of FIG. 10, the branching state determining unit 105 sets the degranulation tag = 1 so as to indicate that there is no wrinkle. Threshing refers to a state in which cocoons are falling from the ears. Thereafter, the process proceeds to step S11190.

  In step S11190 of FIG. 10, the branch state determination unit 105 adds each attribute value to the grain tag. Each attribute value is trait information representing the state of the cocoon, and includes information on the cocoon grain size, the cocoon “ま た は” or “sterile” state, and degenerated buds. The grain size of the cocoon includes length and width. The measurement of the length and width of the buttock was described in the description of step S3020 in FIG. As described above, the “稔” state of the cocoon refers to a state where the fruit is present in the rice husk, and the “non-sterile” state refers to a state where no fruit exists in the husk. The determination of the “稔” state or the “not alive” state has been described in the description of steps S09070 to S09100 in FIG. Degenerate flower is a state where the flower does not bear fruit. The information on the degenerated spikelets may be input by the operator in step S070 of FIG.

  Through the above processing, grain tags or branch tags for all end points are created in the configuration data file 115.

  In step S11200 of FIG. 11, the branch state determination unit 105 starts from the head.

  In step S11210 of FIG. 11, the branch state determination unit 105 uses the configuration data file 115 to determine a route from the starting point to all end points.

  In step S11220 in FIG. 11, the branch state determination unit 105 calculates the distances of all routes using the configuration data file 115.

  In step S11230 of FIG. 11, the branch state determination unit 105 sets the branch point and end point on the longest path including the starting point in the configuration data file 115 as the branch point and end point of the main branch. The main branch including the starting point that is the head is the cob. A branch branch branched from the cob is referred to as a primary branch, a branch branched from the primary branch is referred to as a secondary branch, and a branch branched from the secondary branch is referred to as a tertiary branch. In addition, when the head is designated, the branch connecting the head and the head may be used as the cob.

  In step S11240 in FIG. 11, the branch state determination unit 105 sets a flag of the main branch included in the point tag of the branch point and the end point that are the main branch in the configuration data file 115. The flag of the main branch included in the point tag is set when the branch point or the end point corresponding to the point tag is located on the main branch.

  In step S11250 of FIG. 11, the branch state determination unit 105 determines whether or not there is an unprocessed branch point on the main branch in the configuration data file 115. An unprocessed branch point is a branch point connected to a downstream branch branch whose path distance is not calculated. If there is no unprocessed branch point on the main branch, the process ends. If there is an unprocessed branch point on the main branch, the process proceeds to step S11260.

  In step S11260 of FIG. 11, the branch state determination unit 105 starts from the other branch point connected to the unprocessed branch point, and proceeds to step S11210.

  Here, the configuration of the configuration data file 115 will be described in correspondence with the branching state of the branches.

  FIG. 13 is a diagram illustrating an example of a branching state of branches. In FIG. 13, T1 represents a starting point and T2 represents an end point. D1, D2 and D3 represent branch points.

  FIG. 14 is a diagram showing the configuration of the form data file corresponding to the branch state of the branch shown in FIG. The form data file has a tree structure having point tags corresponding to the start point, end point, and branch point as components. The point tags excluding the starting point tag are child elements of other point tags. In this way, all point tags are related by a parent-child relationship, that is, a hierarchical relationship in which the parent is higher and the child is lower. Each point tag includes information such as a corresponding branch point and branch component information.

  In FIG. 14, specifically, T1 and D1 represent point tags corresponding to the start point T1 and the branch point D1. The point tag T1 has a point tag D1 as a child element, the point tag D1 has point tags D2 and D3 as child elements, and the point tag D3 has point tags T2 and T3 as child elements. Child elements of the point tag D2 are not shown. In FIG. 14, (main) and (minute) represent that the flag of the main branch is set and that the flag is not set, respectively. For example, D2 being (main) indicates that the branch component D1-D2 is the main branch on the basis of D1. On the other hand, since D3 is (minutes), the branch components D1-D3 are branch branches on the basis of D1. Assuming that the branch component T1-D1 is a cob, the branch component D1-D3 is a primary branch, and the branch component D3-T3 is a secondary branch. As described above, the order of the branch infarct including the point in question can be determined based on the number of flags of the main branches that are not standing on the route from the starting point to the point in question.

  FIGS. 18 to 23 are diagrams showing examples of spike diagrams created by the branch state determination unit 105 using only the information of the form data file 115.

  FIG. 18 is a diagram illustrating an example of a panicle diagram using position information of branch components.

  FIG. 19 is a diagram illustrating an example of a spike-shaped diagram schematically used without using position information and length information of branch components. The length of the branch component is set to a predetermined value determined according to the order of the branch infarct.

  FIG. 20 is a diagram showing an example of a schematic spike diagram in which the wrinkles are removed from the spike diagram of FIG. 19 and only the branches are left.

  FIG. 21 is a diagram showing an example of a schematic panicle diagram in which the length information of the branch component is used to remove the wrinkles and leave only the branches.

  FIG. 22 is a diagram illustrating an example of a spike diagram schematically represented three-dimensionally (three-dimensionally) using branch component length information. When creating a three-dimensionally modeled panicle diagram, predetermined fixed values are used for the angle between the branch branch and the main branch, the direction in which the branch branch extends, and the like.

  FIG. 23 is a diagram illustrating an example of a three-dimensionally modeled panicle diagram in which the length information of the branch component is used to remove the wrinkles and leave only the branches.

  FIG. 12 is a flowchart showing the detailed process of step S120 of FIG.

  In step S <b> 12010 of FIG. 12, the data management unit 107 stores the branch point area in the storage file 117.

  In step S <b> 12020 of FIG. 12, the data management unit 107 saves the end point area in the save file 117.

  In step S12030 of FIG. 12, the data management unit 107 saves the branch area in the save file 117.

  In step S12040 of FIG. 12, the data management unit 107 saves the cocoon area in the save file 117.

  In step S12050 of FIG. 12, the data management unit 107 stores the setting parameter in the storage file 117. The setting parameters include upper and lower limits of a predetermined range of the cocoon size and the branch size.

  In step S12060 of FIG. 12, the data management unit 107 determines whether the head position is set. If the head position has been set, the process proceeds to step S12070. If the head position is not set, the process proceeds to step S12080.

  In step S12070 of FIG. 12, the data management unit 107 stores the head position in the storage file 117.

  In step S12080 of FIG. 12, the data management unit 107 determines whether the top position is set. If the head position has been set, the process proceeds to step S12090. If the head position is not set, the process is terminated.

  In step S12090 of FIG. 12, the data management unit 107 stores the top position in the storage file 117.

  By saving the editing state as described above, it can be easily resumed after the work is interrupted.

  According to the above embodiment, the work for digitizing the shape of the ear can be greatly reduced, and the work requiring skill is not necessary. In addition, the length of a curved branch that is difficult to measure manually can be measured. Furthermore, for a panicle that cannot be expanded at once due to multiple grains, it is possible to analyze (measure) by dividing the cob and the primary branch rachis.

  According to the above-described embodiment, the parameters regarding the form of the ear are collected in the form data file 115. Therefore, when analyzing the properties of the morphology of the ear, it is convenient because the configuration data file 115 can be used repeatedly without re-measuring the morphology of the ear. Further, since various two-dimensional and three-dimensional spike diagrams as shown in FIGS. 20 to 23 can be created from the shape data file 115, it becomes easier to grasp the shape of the spike.

  The above embodiments have been described using rice spikelets as an example. However, the analysis method and analysis apparatus of the present invention can also be applied to the ears of grains (plants) other than rice.

It is a figure which shows the structure of the analyzer by embodiment of this invention. 3 is a flowchart illustrating an analysis method according to an embodiment of the present invention. It is a flowchart which shows the detailed process of step S030 of FIG. It is a flowchart which shows the detailed process of step S040 of FIG. It is a flowchart which shows the detailed process of step S050 of FIG. 3 is a flowchart showing a detailed process of step S060 of FIG. It is a flowchart which shows the detailed process of step S090 of FIG. It is a flowchart which shows the detailed process of step S100 of FIG. It is a flowchart which shows the detailed process of step S110 of FIG. It is a flowchart which shows the detailed process of step S110 of FIG. It is a flowchart which shows the detailed process of step S110 of FIG. It is a flowchart which shows the detailed process of step S120 of FIG. It is a figure which shows an example of the branch state of a branch. It is a figure which shows the structure of the form data file corresponding to the branch state of the branch shown by FIG. It is a figure which shows an example of the image of the binarized ear. It is a figure which shows an example of the image of the extracted buttocks. It is a figure which shows an example of the image of the extracted branch part. It is a figure which shows an example of the ear-type figure using the positional information on a branch component. It is a figure which shows an example of the spikelet figure which modeled without using the positional information and length information of a branch component. It is a figure which shows an example of the schematic ear | edge type | mold figure which removed the wrinkles from the ear | edge type | mold figure of FIG. 19, and left only the branch. It is a figure which shows an example of the schematic ear | edge type | mold figure which used the length information of the branch component, removed the wrinkles, and left only the branch. It is a figure which shows an example of the ear | edge type | mold figure stereomorphically represented using the length information of a branch component. It is a figure which shows an example of the three-dimensionally modeled panicle figure which used the length information of the branch component, removed the wrinkles, and left only the branch.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 ... Image collection part, 103 ... Partial extraction part, 105 ... Branch state determination part, 107 ... Data management part, 109 ... Input / output part, 115 ... Form data file

Claims (9)

An analysis method for analyzing a branch structure of a plant ear , executed by a computer ,
Acquiring a panicle image;
Extracting branches and paddy image by processing the image of the ear,
Identifying a branch state of the branch from the extracted branch image ;
Including
Step, the state of a branch of the branches identified, analyzed method comprising in the storage device as a data file having a tree structure corresponding to the state of the branches of the branch to determine the status of the branch of the branch. Identifying a state of branching of the branch defines a first starting point on the branch located on the base of ear, the first is also a branch point of the branch you adjacent to the starting point as an end point to end point, the first from the start point to the end point to the branch component, further, the end point as a new starting point the end point in the case of a branch point, the branch point or end point of the branch adjacent to the starting point as an end point, the from the start point The analysis method according to claim 1, comprising dividing the entire branch of a spike into branch components by repeatedly determining the end point as another branch component . Wherein the data file includes a point tags corresponding to the point tags and end points corresponding to the branch point,
Points tag corresponding to the branch point is provided with a point tag corresponding to the end point of the branch component whose starting point is the branch point as a child element,
It said point tag corresponding to the end point, the length of the information including the branch component from the start point to the end point, the analysis method according to 請 Motomeko 2. Points tag corresponding to the branch point has the information of the branch point and the information of the branch component whose starting point is the branch point,
The analysis method according to claim 3. Points tag corresponding to the end point, has a rice information attached to the said end point, the method of analysis according to claim 3. The rice of information includes trait information representative of the paddy conditions The method of analysis according to claim 5. Identifying a state of branching of the branch, the path formed by said branches component a path connecting between the longest end point and the initial starting distance from the first starting point, the first determined primarily edge corresponding to the starting point, comprising a branch branching from the main branch determined by the branches of the main branch, the method of analysis according to claim 2. Wherein the data file includes a point tags corresponding to the point tags and end points corresponding to the branch point,
Points tag corresponding to the branch point is provided with a point tag corresponding to the end point of the branch component whose starting point is the branch point as a child element,
The point tag corresponding to the end point has information on the branch component from the start point to the end point ,
The branch component information includes information on whether the branch component is a main branch or a branch branch .
The analysis method according to claim 7. An analysis device for analyzing the branching structure of a plant ear,
An image collection unit for acquiring an image of the ear;
A portion extraction unit for extracting a branch and rice image by processing the image of the ear,
A branch state determination unit that identifies a branch state of the branch from the extracted branch image ;
Including
The branch state determination unit stores the branch state of the specified branch in a storage device as a data file having a tree structure corresponding to the branch state of the branch .
Analysis device.

Images