CN115294459B - Wood growth ring identification method based on growth imbalance characteristic - Google Patents
Wood growth ring identification method based on growth imbalance characteristic Download PDFInfo
- Publication number
- CN115294459B CN115294459B CN202211186392.6A CN202211186392A CN115294459B CN 115294459 B CN115294459 B CN 115294459B CN 202211186392 A CN202211186392 A CN 202211186392A CN 115294459 B CN115294459 B CN 115294459B
- Authority
- CN
- China
- Prior art keywords
- area
- sunward
- annual ring
- region
- angle
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/36—Applying a local operator, i.e. means to operate on image points situated in the vicinity of a given point; Non-linear local filtering operations, e.g. median filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Nonlinear Science (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the field of identification methods by using electronic equipment, in particular to a wood growth ring identification method based on the characteristic of unbalanced growth, which comprises the following steps: acquiring an image of a wood annual ring; performing edge detection on the annual ring image, and converting Cartesian coordinates of edge points into polar coordinates to obtain an angle area; clustering edge points in the angle area to obtain the dispersion of the angle area; obtaining a clustering chain by using the polar coordinates of the clustering centers in the angle area, and obtaining a annual ring line by using the clustering chain; acquiring central areas of the sunward area and the back-sun area, and performing iterative combination on an angle area between the two areas to obtain the sunward area and the back-sun area; and carrying out Hough circle detection on the annual ring lines in the sunny region and the back-sunny region, splicing Hough circles belonging to the same annual ring line in the sunny region and the back-sunny region, and identifying the annual ring line in the wood annual ring image. The method is used for wood growth ring identification, and the accuracy of growth ring identification can be improved by the method.
Description
Technical Field
The invention relates to the field of identification methods by using electronic equipment, in particular to a wood growth ring identification method based on the characteristic of unbalanced growth.
Background
The wood annual rings not only provide the age information of one tree but also provide the information of the nature, and researchers can calculate the width of each annual ring to research and analyze the growth condition of the tree, so the wood annual rings have high research value. Annual ring identification is the basis of annual ring analysis, so that annual ring identification is necessary. However, due to the influence of uneven illumination, the tree growth rings are not uniformly distributed, generally, the south side of the tree receives sufficient illumination, more nutrients are obtained, the growing field is thicker, the growth rings are more sparse, the north side of the tree receives less sunlight, and the growth rings are more compact. Therefore, the invention provides a method for identifying a wood annual ring by applying electronic equipment, wherein Hough circle transformation is carried out on different areas of wood, annual ring lines obtained on different areas are combined to obtain a final tree annual ring line, and further, the annual ring area is finely identified.
Disclosure of Invention
The invention provides a wood annual ring identification method based on the characteristic of unbalanced growth, which comprises the following steps: acquiring an image of a wood annual ring; carrying out edge detection on the annual ring image, and converting the Cartesian coordinates of edge points into polar coordinates to obtain an angle area; clustering edge points in the angle area to obtain the dispersion of the angle area; obtaining a clustering chain by utilizing the polar coordinates of a clustering center in the angle area, and obtaining a annual ring line by utilizing the clustering chain; acquiring central areas of the sunward area and the back-sun area, and performing iterative combination on an angle area between the two areas to obtain the sunward area and the back-sun area; compared with the prior art, the tree growth method based on the Hough circle detection has the advantages that wood is divided into sunny areas and sunny back areas according to the density degree of the tree rings based on the unbalanced characteristic of the tree in the growth process, and the edge contours of the tree ring lines are extracted. And then carrying out Hough circle fitting on the annual ring lines in different areas respectively based on the difference of the density degrees of the annual ring lines in the sunward area and the dormitory area, combining the annual ring lines obtained in different areas to obtain the final tree annual ring line, and further realizing finer identification of the annual ring areas.
In order to achieve the purpose, the invention adopts the following technical scheme that the wood growth ring identification method based on the growth imbalance characteristic comprises the following steps:
and acquiring the preprocessed wood annual ring image.
And performing edge detection on the wood annual ring image, converting the acquired Cartesian coordinates of each edge point into polar coordinates, and averagely dividing the value range of the polar angle in the polar coordinates to obtain all angle areas.
And clustering the edge points in each angle area, and calculating the dispersion of each angle area according to the difference value of each clustering center in each angle area on the polar coordinate.
And obtaining all clustering chains by using the polar coordinates of the clustering centers in each angle area.
And combining the edge points contained in each clustering chain to obtain all annual ring lines in the wood annual ring image.
Selecting the angle areas with the largest dispersion and the smallest dispersion from all the angle areas, taking the angle area with the largest dispersion as the central area of the sunward area, and taking the angle area with the smallest dispersion as the central area of the sunward area.
And taking the central area of the sunward area as a starting point, and iteratively combining the angle areas between the central areas of the sunward area and the sunward area to obtain the sunward area and the sunward area.
And performing Hough circle detection on each annual ring line in the sunny region and the sunny region according to the gradient of the edge points on the annual ring lines to obtain Hough circles corresponding to each annual ring line in the sunny region and the sunny region.
And splicing Hough circles belonging to the same annual ring line in the sunny area and the sunny back area, and identifying all complete annual ring lines in the wood annual ring image.
Further, in the wood growth ring identification method based on the growth imbalance characteristic, the dispersion of each angle region is obtained as follows:
and carrying out edge detection on the preprocessed wood annual ring image to obtain the Cartesian coordinates of each edge point in the preprocessed wood annual ring image.
And converting the Cartesian coordinates of each edge point into polar coordinates, and averagely dividing the value range of the polar angle in the polar coordinates to obtain all angle areas.
And performing iterative clustering on the edge points in the first angle area to obtain the number of clustering clusters, and taking the number of clustering clusters as the number of annual ring lines in the wood annual ring image to be identified.
And calculating the dispersion of the first angle area according to the difference value of each clustering center in the first angle area on the polar coordinate.
And clustering the edge points in the remaining angle areas by using the cluster number of the first angle areas, and calculating the dispersion of each remaining angle area according to the difference value of each cluster center in each angle area on the polar coordinate.
Further, in the wood growth ring identification method based on the growth imbalance characteristic, all the growth ring lines in the wood growth ring image are obtained as follows:
and sequencing and numbering the clustering centers in each angle area according to a mode that the polar coordinates are from small to large.
And sequencing the clustering centers with the same number in each angle area according to the angle range from small to large to obtain a clustering chain corresponding to each annual ring line.
And combining the edge points contained in the clustering chain corresponding to each annual ring line to obtain all the annual ring lines in the wood annual ring image.
Further, in the wood growth ring identification method based on the growth imbalance characteristic, the sunny region and the back-sunny region are obtained as follows:
and selecting the angle areas with the maximum dispersion and the minimum dispersion from all the angle areas, taking the angle area with the maximum dispersion as the central area of the sunny area, and taking the angle area with the minimum dispersion as the central area of the dormitory area.
Marking one side of a spacing region between a central region of a sunward region and a central region of a sunward region as
Side, the other side is marked as
And (3) side.
With the central area of the sunward area as a starting point, all the remaining angle areas are firstly allocated to the sunward area, so as to obtain a first sunward area and a first sunward area.
And calculating the dispersion mean value and the dispersion variance in the class of the first sunward area and the first sunward area.
And calculating to obtain the first sunward-backsun region area division by utilizing the number of angle regions, the dispersion mean value and the intra-class dispersion variance contained in the first sunward region and the first backsun region.
Placing the first sunward region in
And combining an angle area laterally and forwards to obtain a second sunward area and a second dorsiform area.
And calculating to obtain the second sunward-backsun region area division by utilizing the number of angle regions, the dispersion mean value and the intra-class dispersion variance contained in the second sunward region and the second backsun region.
Judging the area division of the second sunny-sunny area: when the second sunward-backsun region is higher than the first sunward-backsun region, the second sunward region is used as the updated sunward region and continues to be
Combining laterally and forwards, when the second sunward-dorsolarium region is lower than the first sunward-dorsolarium region, the first sunward region is
The merging of the sides stops and the first sunward region is taken as the updated sunward region
Merging laterally and anteriorly.
The updated sunward area is arranged in
Side and
the side is iteratively combined until the updated sunward area is in
Side and
and when the side can not continue to merge the angle areas, stopping iterative merging to obtain the sunward area and the dormitory area.
Further, in the wood annual ring identification method based on the growth imbalance characteristic, the expression of the sunny-sunny region division degree is as follows:
in the formula (I), the compound is shown in the specification,
for the division of the sunny-dorsiflexion region,
and
the variance of the dispersion in class of the sunny area and the back-sunny area respectively,
and
the number of the angle regions included in the sunny region and the dorsiflexion region,
and
the mean of the dispersion of the sunny and sunward areas is shown,
represent
And
the greater the value of the number of the first,
and
representing the weight coefficients.
Further, according to the wood growth ring identification method based on the growth imbalance characteristic, all complete growth ring lines in the wood growth ring image are identified as follows:
and carrying out Hough circle detection on each annual ring line contained in the sunny region according to the gradient of the edge points on the annual ring lines to obtain the circle center position and the radius of each annual ring line in the sunny region.
And obtaining the circle center position and the radius of each annual ring line in the sunny region according to the mode of obtaining the circle center position and the radius of each annual ring line in the sunny region.
And obtaining Hough circles of each annual ring line in the sunward area and the dorsolateral area according to the circle center position and the radius of each annual ring line in the sunward area and the dorsolateral area.
And splicing Hough circles belonging to the same annual ring line in the sunny area and the sunny back area, and identifying all complete annual ring lines in the wood annual ring image.
Further, in the wood growth ring identification method based on the growth imbalance characteristic, the preprocessed wood growth ring image is obtained as follows:
and collecting an image of the annual ring of the wood to be identified.
And carrying out gray processing on the wood annual ring image to be identified to obtain a wood annual ring gray image.
And denoising the wood annual ring gray map to obtain a denoised wood annual ring gray map.
And enhancing the denoised wood annual ring gray level image to obtain a preprocessed wood annual ring image.
The invention has the beneficial effects that:
according to the method, based on the unbalanced characteristic of the tree in the growing process, the wood is divided into a sunny area and a back-sunny area according to the density degree of the annual rings, and the edge outline of the annual ring line is extracted. And then carrying out Hough circle fitting on the annual ring lines in different areas respectively based on the difference of the density degrees of the annual ring lines in the sunward area and the dormitory area, combining the annual ring lines obtained in different areas to obtain the final tree annual ring line, and further realizing finer identification of the annual ring areas.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a wood growth ring identification method provided in embodiment 1 of the present invention;
fig. 2 is a schematic flow chart of a wood growth ring identification method according to embodiment 2 of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
Example 1
The embodiment of the invention provides a wood annual ring identification method based on the characteristic of unbalanced growth, which comprises the following steps of:
and S101, acquiring the preprocessed wood annual ring image.
The preprocessing comprises graying, denoising and image enhancement.
S102, carrying out edge detection on the wood annual ring image, converting the acquired Cartesian coordinates of each edge point into polar coordinates, and averagely dividing the value range of the polar angle in the polar coordinates to obtain all angle areas.
Wherein the edge detection employs
And (5) an operator.
S103, clustering the edge points in each angle area, and calculating the dispersion of each angle area according to the difference value of each clustering center in each angle area on the polar coordinates.
In clustering, a data set is identified into different classes or clusters according to a certain criterion (e.g., a distance criterion), so that the similarity of data objects in the same cluster is as large as possible, and the difference of data objects not in the same cluster is also as large as possible. Namely, after clustering, the data of the same class are gathered together as much as possible, and different data are separated as much as possible.
And S104, obtaining all clustering chains by using the polar coordinates of the clustering centers in each angle area.
And sorting and numbering the clustering centers by using the size of the polar coordinates.
And S105, combining the edge points contained in each clustering chain to obtain all annual ring lines in the wood annual ring image.
Wherein each clustering chain corresponds to one annual ring line.
S106, selecting the angle areas with the maximum dispersion and the minimum dispersion in all the angle areas, taking the angle area with the maximum dispersion as the central area of the sunny area, and taking the angle area with the minimum dispersion as the central area of the dormitory area.
Wherein the dispersion reflects the degree of density.
And S107, iteratively combining the angle areas between the central areas of the sunward area and the back sun area by taking the central area of the sunward area as a starting point to obtain the sunward area and the back sun area.
And judging whether to continue merging or not by calculating the distinguishing degree of the sunny-sunny region after merging every time.
And S108, carrying out Hough circle detection on each annual ring line in the sunny region and the back-sunny region according to the gradient of the edge points on the annual ring lines to obtain a Hough circle corresponding to each annual ring line in the sunny region and the back-sunny region.
Among them, hough transform is a very important method for detecting the shape of the boundary of a discontinuity. The method realizes the fitting of a straight line and a curve by transforming the image coordinate space to the parameter space.
And S109, splicing Hough circles belonging to the same annual ring line in the sunny region and the sunny region, and identifying all complete annual ring lines in the wood annual ring image.
The Hough circles in different areas are spliced to fit a more practical annual ring line.
The beneficial effect of this embodiment lies in:
the embodiment divides the wood into a sunny region and a back-sunny region according to the density degree of the annual rings based on the unbalanced characteristic of the trees in the growing process, and extracts the annual ring line edge profile. And then carrying out Hough circle fitting on the annual ring lines in different areas respectively based on the difference of the density degrees of the annual ring lines in the sunward area and the dormitory area, combining the annual ring lines obtained in different areas to obtain the final tree annual ring line, and further realizing finer identification of the annual ring areas.
Example 2
The embodiment of the invention provides a wood growth ring identification method based on growth imbalance characteristics, as shown in fig. 2, comprising the following steps:
s201, carrying out image preprocessing on the annual ring image.
The annual ring image is subjected to image preprocessing to be
The color image is converted to a grayscale image and then filters the noise in the image using filters such as gaussian, median and mean filters. The gamma transformation is adopted to carry out enhancement operation on the image so as to weaken the whole image caused by too dark or too bright imaging environment in the image acquisition processInfluence of volume gray scale.
And S202, dividing an angle area.
Due to the influence of uneven illumination, the annual rings of the trees are not uniformly distributed, generally, the south side of the trees receives sufficient illumination, more nutrients are obtained, the growing field is thicker, the annual rings are sparser, the north side of the trees receives less sunlight, and the annual rings are tighter. The circular contour of tree growth rings cannot be well fitted by adopting the conventional Hough circle transform.
First adopt
Operator extraction of wood section image
Using a Gaussian filter to smooth the image
And to
And (3) obtaining the gradient amplitude and the gradient direction of each pixel, judging the gradient amplitude of each pixel and the gradient amplitude of the adjacent pixel in the gradient direction, carrying out non-maximum value inhibition, and detecting edges by adopting a dual-threshold algorithm and connecting the edges. Because the shape of the annual ring line is mostly a closed curve approximate to a circle, the distances between points on the same annual ring line and the center are approximately equal, the calculation amount is reduced (only the distance between the points and the center is considered) and the sunny area and the back-sunny area are conveniently divided (the circular area is divided into two fan-shaped areas), and the coordinate information of the edge points is obtained
Conversion from Cartesian to polar coordinates
Wherein
，
Will be
Is divided into
Angular region, divided into
（
），
（
），
（
），
，
（
）。
And S203, obtaining the dispersion of each angle area.
Are respectively at
Polar coordinates of edge points on inner annual ring line
Clustering operation is carried out, edge points belonging to different annual ring lines are layered, the average distance between different layers is calculated to judge the density degree of annual ring distribution, and then angle areas are respectively judged
The dispersion of (2). The specific method is to
All edge points in the angular region
Performing clustering operation, selecting
（
From
Starting to select) edge points as
Initial clustering center of each class, calculating its arrival at each edge point
Distance of each cluster center and assigning them to the minimum distance
In a class, then to this
Class calculates the centroid of all samples as new
In clustering of individual classesHeart until one of the termination conditions is met: 1. the number of iterations is
；2.
The cluster centers of the individual classes no longer change.
Is obtained by
After class classification, evaluation is carried out according to the distribution situation of edge point coordinates in the class
The clustering effect of the individual classes is such that,
the closer the number of the annual rings is to the circle number of the annual rings, the more concentrated the distribution of the edge points in the class is, the better the clustering effect is, and the calculation is adopted
Clustering evaluation criterion of individual classes
. Wherein
Denotes the first
The number of edge points in the individual class,
indicating the second within the angular region
The first in a class
The number of the edge points is equal to or less than the number of the edge points,
denotes the first
The center of the cluster in the individual class,
representing edge points
And cluster central point
In polar coordinates
The difference in the above. Then increase
Is obtained by repeating the above steps
Clustering, calculating the current
Clustering evaluation criterion under value
Generally, the number of annual rings does not exceed 20 (the implementer can design according to the maximum growth cycle of the tree), and the maximum is set
Is taken as
. According to cluster evaluation criteria
Find the most suitable
The number of layers to be classified and the number of annual ring lines are calculated
Time of flight
Dispersion of individual cluster centers
As
The dispersion of the intervals.
Generally speaking, the annual rings are distributed in different angle areas although the density degree of the parts in different orientations is different
Is consistent in, and therefore adopts
Continue to the rest area
Performing clustering operation to obtain
Corresponding dispersion
。
And S204, acquiring a annual ring line.
Then the angle area is divided into
Is/are as follows
Are combined to form the edge points in the class
Annual ring lines. The specific way is to angle area
In algorithms for pairwise matching
Cluster center point of class (common)
Individual cluster center points) because the initial center points during clustering are randomly selected during clustering, and the cluster centers labeled with the same cluster number in different angle intervals do not necessarily belong to the same annual line in polar coordinate space, i.e., the cluster centers are randomly selected
Is not necessarily equal to
The first type of (2) is on the same annual ring line, so it is necessary to find
Clustering chain of annual ring lines
Such as
Represent
Annual ring lines are respectively divided into angle areas
After clustering of
And (4) class composition. Considering that the annual ring lines generally do not cross, the annual ring lines can be firstly aligned
In (1)
Clustering center point of class according to polar coordinate
Ordered from small to large, and according to
Sorting of center points of middle clusters by polar coordinates
The cluster number of the minimum cluster central point forms a cluster chain
Polar coordinates
The cluster numbers of the second smallest cluster center point form a cluster chain
，
Polar coordinates
The cluster numbers of the maximum cluster central points form cluster chains
. Finding a clustering chain
Then will be
In (1)
Clustering the set of edge points of a class according to a cluster chain
Are combined into a complete annual ring line
。
And S205, dividing the angle area into a sunny area and a backsun area.
Then according to the obtained angle region
Corresponding dispersion
The difference is that the angle area is divided into a sunny area and a sunny-back area, and the sunny area and the sunny-back area do not always occupy the section of the wood respectively because the trees receive the light which is not uniform
Half of so according to the angular area
The distribution condition of the dispersion degree divides the area of the trees facing the sun and the area of the trees facing the sun. First, the angular region is taken
The area with the maximum and minimum dispersion is used as the sunward area
And the area of the back and sun
To the remaining central region of
The angular regions are divided, and due to the uniformity and regularity of tree growth, generally speaking, the regions with the maximum dispersion and the minimum dispersion are not closely adjacent, and are provided with spaced regions, and the sunward regions are arranged
And the area of the back and sun
Is spaced apart from the central region by
The other side is
And (3) side. Starting from the central region of the sunward region, first, the sun-ward region is divided into two regions
All the angle areas are distributed to the area of the back sun, and the area of the back sun at the moment is calculated
Mean value of dispersion of
And variance of intra-class dispersion
(the mean value of the dispersion of the sunward region is the sunward region at this time
Initial dispersion corresponding to the central region of (1)
Variance of initial intra-class dispersion
Is composed of
) At this time
In that
Swallowed sideways and forward by an angular zone, respectively
In that
Laterally retracting an angle area, respectively calculating the current time
And
mean value of dispersion of
And variance of intra-class dispersion
If it is present
And
the smaller the intra-class dispersion variance sum of (c),
and
the larger the dispersion mean difference value of (3), the division is illustrated
And
the better the effect of (A), thereby designing the division of the sunny-sunny region
Comprises the following steps:
wherein
And
are respectively the current
And
the variance of the intra-class dispersion of (c),
and
is at present
And
angle included inThe number of the regions is such that,
and
to represent
And
the mean value of the dispersion of (a),
represent
And
the greater the value of the number of the first,
and
representing weight coefficients, setting
，
. Division of sunny-sunny back region
The larger the size, the sunny region is indicated
And the area of the back and the sun
The greater the dispersion of (a) is distinguished,partitioning
And
the better the effect of the two regions.
Then updated
And
area coverage, then let
In that
Advancing a region laterally and forwards, and solving the division of the sunward-dorsolateral region after swallowing
Comparing the current
Before value and swallow
The magnitude of the value, if
Increasing the size of the area, swallowing the area in the sun, stopping swallowing the area if the size of the area is not increased, and repeating the operation until the size of the area in the sun is increased
Side and
the side can not continue to swallow the region position, and the sunny region is obtained
And the area of the back and sun
(
And
the number of the angular areas occupied by the areas is respectively
And
). Will be exposed to the sun
And the area of the back and the sun
As the identification line
And
will be exposed to the sun
And the area of the back and the sun
Distinguishing and obtaining an identification line
And
coordinates of edge points of (a), (b)
And
respectively represent identification lines
And
go to the first
Edge points on the annual ring line of the tree, wherein
Is in the value range of
）。
And S206, identifying the annual ring area.
According to the analysis, the trees can be found in the sunny area under the influence of the irradiation direction of the sunlight
And the area of the back and sun
The growth of the tree is not balanced, the sunny area grows robustly, the annual ring lines are sparse, the radius of a fitted Hough circle is larger on the same annual ring line, the annual ring line of the sunny area is tight, and the radius of the fitted Hough circle is smaller, so that the contour information of the tree annual ring line cannot be completely fitted by adopting the conventional Hough circle transformation, and the tree annual ring line is respectively aligned to the sunny area
And the area of the back and the sun
Fitting Hough circle, firstly, to the sunny region
Identification line of
Point on the first annual ring line of the upper edge point
Beginning at the first annual ring line
Go up to
Calculating the gradient of edge and determining circumference line, the gradient of point on the circumference is the normal of circle, drawing gradient straight line of all figures on polar coordinate, the larger the value of summation on coordinate point is, the more times the straight line is crossed on the point is, the more likely it is the center of circle, carrying out non-maximum value suppression in 4 neighborhoods of Hough space, determining in sunny region
The position of the center of a circle of the upper first annual ring line
And radius
. The same method is adopted for the sunward areas
On
Carrying out Hough semicircle detection to respectively obtain the circle center position
And radius
. Then to the back sun region
Identification line of
Point on the first annual ring line of the upper edge point
Beginning at the first annual ring line
Go up to
So far, the same procedure is adopted to find the area of the sun-back
Position of center of circle on
And radius
. Finally, the sunward area is formed
And the area of the back and sun
The obtained Hough circles are merged, and the annual ring line is obtained
Will be the same annual ring line
（
From
Get to
) Obtained as above
And
different Hough circles of the region
And
and (4) splicing (splicing of two intersected circles) to form a complete annual ring line, and the step of identifying the annual ring area is completed.
The beneficial effect of this embodiment lies in:
the embodiment divides the wood into a sunny region and a back-sunny region according to the density degree of the annual rings based on the unbalanced characteristic of the trees in the growing process, and extracts the annual ring line edge profile. And then, on the basis of the difference of the density degrees of the annual ring lines of the sunward area and the sunward area, hough circle fitting is respectively carried out on the annual ring lines in different areas, the annual ring lines obtained in different areas are combined to obtain the final tree annual ring line, and the annual ring areas are identified more finely.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (4)
1. A wood growth ring identification method based on the characteristic of unbalanced growth is characterized by comprising the following steps:
acquiring a preprocessed wood annual ring image;
performing edge detection on the wood annual ring image, converting the acquired Cartesian coordinates of each edge point into polar coordinates, and averagely dividing the value range of the polar angle in the polar coordinates to obtain all angle areas;
clustering the edge points in each angle area, and calculating the dispersion of each angle area according to the difference value of each clustering center in each angle area on the polar coordinate;
the dispersion of each angle region is obtained as follows:
performing edge detection on the preprocessed wood annual ring images to obtain Cartesian coordinates of each edge point in the preprocessed wood annual ring images;
converting the Cartesian coordinates of each edge point into polar coordinates, and averagely dividing the value range of the polar angle in the polar coordinates to obtain all angle areas;
performing iterative clustering on the edge points in the first angle area to obtain the number of clustering clusters, and taking the number of clustering clusters as the number of annual ring lines in the wood annual ring image to be identified;
calculating the dispersion of the first angle area according to the difference value of each clustering center in the first angle area on the polar coordinate;
clustering the edge points in the remaining angle areas by using the number of the clustering clusters in the first angle area, and calculating the dispersion of each remaining angle area according to the difference value of each clustering center in each angle area on the polar coordinate;
obtaining all clustering chains by using the polar coordinates of the clustering centers in each angle area;
sorting and numbering the clustering centers in each angle area according to a polar coordinate from small to large;
sorting the clustering centers with the same number in each angle area according to a mode that the angle range is from small to large to obtain a clustering chain corresponding to each annual ring line;
combining edge points contained in each clustering chain to obtain all annual ring lines in the wood annual ring image;
selecting angle areas with the maximum dispersion and the minimum dispersion from all the angle areas, taking the angle area with the maximum dispersion as a central area of the sunward area, and taking the angle area with the minimum dispersion as a central area of the sunward area;
taking the central area of the sunward area as a starting point, and iteratively combining the angle areas between the central areas of the sunward area and the sunward area to obtain the sunward area and the sunward area;
the sunward area and the dormitory area are obtained as follows:
marking one side of a spacing region between a central region of a sunward region and a central region of a sunward region as
Side, the other side marked as
A side;
taking the central area of the sunward area as a starting point, firstly, distributing all the remaining angle areas to the sunward area to obtain a first sunward area and a first sunward area;
calculating the dispersion mean value and the dispersion variance in the class of the first sunward area and the first dormitory area;
calculating to obtain a first sunward-backsun region area degree by utilizing the number of angle regions, the dispersion mean value and the intra-class dispersion variance contained in the first sunward region and the first backsun region;
placing the first sunward region in
Combining an angle area in a lateral forward mode to obtain a second sunward area and a second backsun area;
calculating to obtain a second sunward-dormitory region area degree by utilizing the number of angle regions, the dispersion mean value and the intra-class dispersion variance of the second sunward region and the second dormitory region;
judging the second sunny-sunny region discrimination: when the second sunward-backsun region is higher than the first sunward-backsun region, the second sunward region is used as the updated sunward region and continues to be
Combining laterally and forwards, when the second sunward-dorsolarium region is lower than the first sunward-dorsolarium region, the first sunward region is
The merging of sides stops and the first sunward region is taken as the updated sunward region
Merging in a lateral and front mode;
the updated sunward area is arranged in the position
Side and
the side is iteratively combined until the updated sunward area is in
Side and
when the side can not continue to merge the angle areas, the iterative merging is stopped to obtain a sunny area and a back-sunny area;
performing Hough circle detection on each annual ring line in the sunny region and the sunny region according to the gradient of the edge points on the annual ring lines to obtain Hough circles corresponding to each annual ring line in the sunny region and the sunny region;
and splicing Hough circles belonging to the same annual ring line in the sunny area and the sunny back area, and identifying all complete annual ring lines in the wood annual ring image.
2. The wood growth cycle identification method based on the growth imbalance characteristic as claimed in claim 1, wherein the expression of the sunny-sunny region division is as follows:
in the formula (I), the compound is shown in the specification,
for the division of the sunny-dorsiflexion region,
and
the variance of the dispersion in class of the sunny area and the back-sunny area respectively,
and
the number of the angle regions included in the sunny region and the dorsiflexion region,
and
the mean value of the dispersion of the sunny area and the back-sunny area is represented,
to represent
And
the greater the value of the number of the first,
and
representing the weight coefficients.
3. The wood growth imbalance characteristic-based wood growth ring identification method according to claim 1, wherein all complete growth ring lines in the wood growth ring image are identified as follows:
carrying out Hough circle detection on each annual ring line contained in the sunward area according to the gradient of the edge points on the annual ring lines to obtain the circle center position and the radius of each annual ring line in the sunward area;
obtaining the circle center position and the radius of each annual ring line in the sunny region in a mode of obtaining the circle center position and the radius of each annual ring line in the sunny region;
obtaining Hough circles of each annual ring line in the sunward area and the dormitory area according to the circle center position and the radius of each annual ring line in the sunward area and the dormitory area;
and splicing Hough circles belonging to the same annual ring line in the sunny area and the sunny back area, and identifying all complete annual ring lines in the wood annual ring image.
4. The wood growth imbalance characteristic-based wood growth ring identification method according to claim 1, wherein the preprocessed wood growth ring image is obtained as follows:
collecting an image of a wood annual ring to be identified;
carrying out gray processing on the wood annual ring image to be identified to obtain a wood annual ring gray image;
denoising the wood annual ring gray level image to obtain a denoised wood annual ring gray level image;
and enhancing the denoised wood annual ring gray level image to obtain a preprocessed wood annual ring image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211186392.6A CN115294459B (en) | 2022-09-28 | 2022-09-28 | Wood growth ring identification method based on growth imbalance characteristic |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211186392.6A CN115294459B (en) | 2022-09-28 | 2022-09-28 | Wood growth ring identification method based on growth imbalance characteristic |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115294459A CN115294459A (en) | 2022-11-04 |
| CN115294459B true CN115294459B (en) | 2023-01-13 |
Family
ID=83834328
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211186392.6A Active CN115294459B (en) | 2022-09-28 | 2022-09-28 | Wood growth ring identification method based on growth imbalance characteristic |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115294459B (en) |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111815542B (en) * | 2020-07-31 | 2024-05-07 | 南京信息工程大学 | Tree annual ring image medulla positioning and annual ring measuring method |
| CN114677365B (en) * | 2022-04-18 | 2024-04-05 | 北京林业大学 | High-precision tree annual ring analysis method and system |
| CN115063603B (en) * | 2022-07-27 | 2023-04-07 | 海门王巢家具制造有限公司 | Wood annual ring line accurate extraction and restoration method based on edge information |
-
2022
- 2022-09-28 CN CN202211186392.6A patent/CN115294459B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN115294459A (en) | 2022-11-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bao et al. | Field-based architectural traits characterisation of maize plant using time-of-flight 3D imaging | |
| Xiang et al. | Recognition of clustered tomatoes based on binocular stereo vision | |
| CN102324032B (en) | Texture feature extraction method for gray level co-occurrence matrix in polar coordinate system | |
| Sun et al. | Image processing algorithms for infield single cotton boll counting and yield prediction | |
| CN106340016B (en) | A kind of DNA quantitative analysis method based on microcytoscope image | |
| CN115018828A (en) | Defect detection method for electronic component | |
| CN109584281B (en) | Overlapping particle layering counting method based on color image and depth image | |
| Ji et al. | Recognition method of green pepper in greenhouse based on least-squares support vector machine optimized by the improved particle swarm optimization | |
| CN108376403B (en) | Grid colony image segmentation method based on Hough circle transformation | |
| CN101526994B (en) | Fingerprint image segmentation method irrelevant to collecting device | |
| Wang et al. | Recognition and localization of occluded apples using K-means clustering algorithm and convex hull theory: a comparison | |
| CN111882561A (en) | Cancer cell identification and diagnosis system | |
| CN103177257A (en) | Image identification method and image classification method for coprinus comatus | |
| Khedaskar et al. | A survey of image processing and identification techniques | |
| CN111597930A (en) | Coastline extraction method based on remote sensing cloud platform | |
| CN111291686B (en) | Extraction method and system for crop root-fruit phenotype parameters and root-fruit phenotype discrimination method and system | |
| CN105976385A (en) | Image segmentation method based on image data field | |
| CN107977976A (en) | A kind of image partition method | |
| CN115294459B (en) | Wood growth ring identification method based on growth imbalance characteristic | |
| CN113256591A (en) | Device and method for rapidly detecting defects of wide glass | |
| CN104657983B (en) | A kind of fabric ball top Density Detection method filtered based on Gabor | |
| CN104881652B (en) | A kind of line number automatic testing method based on corncob male features | |
| Liu et al. | 3D neuron tip detection in volumetric microscopy images | |
| CN115601543A (en) | Mushroom cluster contour segmentation and reconstruction method based on improved SOLOV2 | |
| CN113408370B (en) | Forest change remote sensing detection method based on adaptive parameter genetic algorithm |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |