A method for extracting a water depth data boundary of a channel

Abstract

The application provides a channel water depth data boundary extraction method. Through preprocessing of water depth data, abnormal points are removed, spatial three-dimensional xyz water depth points are projected and mapped to an xy plane, boundary extraction is started from a point with a minimum value Y, a K-nearest neighbor algorithm is used to obtain candidate points on a boundary line, a next boundary point is determined according to angle sorting of the candidate points, and the cycle is repeated until the starting point A is returned, and some special situations existing in channel water depth data are considered in the extraction process, so that the boundary of arbitrary channel water depth data can be automatically extracted by using the extraction method without prior knowledge of the water depth data.

Description

TECHNICAL FIELD

[0001] The present application relates to discrete point set contour extraction, in particular to a channel water depth data boundary extraction method. BACKGROUND

[0002] Channel water depth data has important reference value for channel design and dredging operation, and using channel water depth data to construct channel digital terrain model can intuitively reflect the underwater topography, which helps people accurately understand the underwater topography and guide channel design and dredging construction. Common digital terrain models include triangular mesh model, regular grid model and hybrid model. However, the water depth points of channel water depth data are generally different in spacing, and the triangular mesh model is suitable for establishing digital terrain model. However, the commonly used triangular mesh model construction method has the convex hull feature, which causes the generation of redundant triangles in the concave area of the digital terrain model generated from discrete points, thereby affecting the accurate expression of channel characteristics by the digital terrain model. The boundary of channel water depth data can accurately express the spatial range covered by the water depth data, and by extracting the boundary of water depth data, the redundant triangles outside the boundary can be removed to realize the accurate construction of triangular mesh model.

[0003] At present, the boundary extraction of channel water depth data is as follows:

[0004] The first idea is to first construct a Delaunay triangular mesh according to the discrete point water depth data, and then delete the edges exceeding the length limit from the outermost layer of the triangular mesh by combining the topological relationship between the points, edges and triangles in the triangular mesh and the edge length threshold, and finally extract the boundary of the water depth data. This method is low in efficiency, needs to construct a triangular mesh first, and the edge length threshold of different channel water depth data is not the same, so it cannot adapt to the automatic extraction of the boundary of any channel water depth data. For related content, please refer to: Jia Juntao et al. in Ocean Surveying and Charting, Vol. 32, No. 2, March 2012, "Research on Key Technologies of Intelligent Extraction of Water Depth Survey Data Boundary"; and Ji Lan et al. in Engineering Graphics Journal, No. 4, 2009, "Improvement of Channel Engineering Water Depth Data Triangulation and Boundary Generation Algorithm".

[0005] The second idea is the rolling ball method, which is based on the alpha-shape algorithm to extract the boundary of point set contour. The construction process can be imagined as a ball with a radius of α rolling on the point set, and the points swept by the ball form a connection to form a surface. Since the size of α is the key to accurately extracting the boundary, too large α will affect the accuracy of the extracted boundary, and too small α will not guarantee the integrity of the boundary. However, the measurement accuracy of actual channel water depth data is not the same, and this method cannot adapt to the automatic extraction of the boundary of any water depth data, and α needs to be determined according to the prior knowledge of the data before the boundary can be accurately extracted. SUMMARY

[0006] The application can overcome the defects of the prior channel water depth data boundary extraction method, can completely automatically extract the boundary of the channel water depth data, does not need to understand the prior knowledge of the data, and has high efficiency in extracting the boundary, and can quickly extract the boundary of large data amount of channel water depth data.

[0007] To achieve the above-mentioned target, the main technical scheme comprises:

[0008] A channel water depth data boundary extraction method, characterized in that channel water depth data is acquired, and the following algorithm process is executed:

[0009] (1) preprocessing of channel water depth data

[0010] The channel water depth data is processed to eliminate repeated points and abnormal points. The channel water depth data is projected and mapped from three dimensions to two dimensions XY plane;

[0011] (2) finding N points closest to the current point, efficiently finding N points closest to the current point according to the K-nearest neighbor method;

[0012] (3) edge intersection test, judging whether two edges intersect;

[0013] (4) calculating the angle between the line connecting the current point and the candidate point and the known edge;

[0014] (5) judging whether the point is inside the polygon.

[0015] The application has the advantages that:

[0016] The method realizes completely automatic extraction of the boundary of the channel water depth data, does not need prior knowledge of the data, significantly improves the extraction speed of the boundary of the channel water depth data, promotes the application of the channel water depth data, and lays a foundation for solving the problems such as redundant grid in constructing a triangular mesh model of the channel water depth data. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 The application is a channel water depth data structure;

[0018] Figure 2 The application is a channel water depth data boundary extraction method flowchart;

[0019] Figure 3 The application is a channel water depth discrete point space form;

[0020] Figure 4 The application is an automatically extracted boundary;

[0021] Figure 5 The application is a digital terrain model redundant grid phenomenon;

[0022] Figure 6 The digital terrain model based on the boundary processing is used for the embodiment of the present application; DETAILED DESCRIPTION

[0023] As Figure 1 shown, the method for extracting the channel water depth data boundary of the present application comprises the following steps:

[0024] The channel water depth data is obtained by measuring the water depth of the channel by a multi-beam water depth measuring instrument, and is composed of a large number of discrete water depth points. Each row in the data content represents a water depth point, which is composed of three values of X, Y and Z. The two values of X and Y represent the plane position, and the value of Z represents the water depth at the current position. For details, see Figure 1 shown.

[0025] (1) The channel water depth data is preprocessed to remove duplicate points and abnormal points, and the channel water depth data is projected and mapped from three dimensions to two dimensions XY plane to form a plane point set P. If the number of points in the point set is less than 3, it is returned without boundary to be generated; if the number of points in the point set is 3, the boundary is formed by all points in the point set; if the number of points is greater than 3, the following steps are continued;

[0026] (2) The point A with the minimum value Y is taken as the starting point of boundary extraction and added to the boundary line point set Q. If there are multiple minimum values, the point with the minimum X value is taken as the starting point, and the point A is removed from the point set P;

[0027] (3) The N points closest to the point A are found as candidate points for the second boundary point. By sorting the angle between the candidate points and the line connecting A and the horizontal X axis, the candidate point with the maximum angle is taken as the second boundary point B. B is added to the boundary line point set Q, and B is removed from the point set P;

[0028] (4) The N points closest to the point B are found as candidate points for the third boundary point. By sorting the angle between the candidate points and the line connecting B and BA, the candidate point with the maximum angle is taken as the third boundary point C. C is added to the boundary line point set Q, and C is removed from the point set P;

[0029] (5) The N points closest to the point C are found as candidate points for the fourth boundary point. By sorting the angle between the candidate points and the line connecting C and CB, the candidate point with the maximum angle which does not intersect with the known edge is taken as the fourth boundary point D. D is added to the boundary line point set Q, and D is removed from the point set P. In addition, if A is not in the point set P, A is added to the point set P;

[0030] (6) The step 5 is repeated until the starting point A is reached, and the boundary line point set Q is obtained;

[0031] (7) judge whether all points except the set Q are in the polygon constituted by the boundary point set Q, if yes, Q is the set of points on the final extracted boundary line, if not, N+1, and repeat steps 3, 4, 5 and 6.

[0032] The embodiment is shown as follows: Figures 3-6 The embodiment is shown as follows:

[0033] Figure 3 The figure is a visualization effect of discrete water depth points in water depth data of a channel, and can directly show the spatial distribution of the water depth points;

[0034] Figure 4 The figure is a boundary obtained by processing the water depth data of the channel by using the channel water depth boundary extraction method and automatically extracting.

[0035] Figure 5 The figure shows the problem of redundant grids formed in the concave hull area by displaying the digital terrain model in the form of a triangular mesh constructed by the channel water depth data;

[0036] Figure 6 The figure is the effect of the digital terrain model after the boundary is automatically extracted and the redundant grids outside the boundary are removed based on the application.

[0037] The application lays a foundation for solving the problem of redundant grids possibly appearing in the construction of a triangular mesh model of channel water depth data by extracting the boundary of the channel water depth data, and based on the boundary, the redundant triangles outside the boundary can be removed, so that the accurate construction of the triangular mesh model is realized, and the application of the channel water depth data is promoted.

Claims

1. A method for extracting a water depth data boundary of a waterway, characterized by, Channel depth data is acquired and the following algorithm process is performed: Step one, preprocessing of channel depth data: The channel depth data is processed to eliminate duplicate points and abnormal points; The channel depth data is projected and mapped from three dimensions to two-dimensional XY plane to form a point set P; Step two, in the point set P described in step one, find the nearest N points from the current point, and efficiently find the nearest N points from the current point according to the K-nearest neighbor method; Further, the intersection test of the edge judges whether two edges intersect or not; Further, the angle between the line connecting the current point and the candidate point and the known edge is calculated; Further, it is judged whether the point is inside the polygon or not; The step two is specifically: (2) Obtain the point A with the minimum value Y as the starting point of boundary extraction, and add it to the boundary line point set Q; (3) Find the nearest N points from point A as candidate points of the second boundary point, determine the second boundary point B by sorting the angle between the line connecting the candidate point and A and the horizontal X axis, and add it to the boundary line point set Q; (4) Find the nearest N points from point B as candidate points of the third boundary point, determine the third boundary point C by sorting the angle between the line connecting the candidate point and B and the BA line, and add it to the boundary line point set Q; (5) Find the nearest N points from point C as candidate points of the fourth boundary point, determine the fourth boundary point D by sorting the angle between the line connecting the candidate point and C and the CB line, and add it to the boundary line point set Q; (6) Repeat step (5) until the starting point A is reached, and obtain the boundary line point set Q; (7) Judge whether all points except the set Q are inside the polygon formed by the boundary point set Q or not, if yes, then Q is the final extracted boundary line point set; if not, then N+1, and repeat steps (3), (4) and (5).

2. The method for extracting the boundary of waterway depth data as described in claim 1, characterized in that, In step one, the data needs to be preprocessed before projection and mapping of channel depth data, and duplicate and abnormal point data is removed. If the number of points is less than 3, the boundary cannot be generated; if the number of points is 3, return the boundary formed by the three points.

3. The method for extracting the boundary of channel depth data as described in claim 1, characterized in that, In step (2), point A needs to be removed from point set P.

4. The method of claim 1, wherein the step of extracting the boundary of the water depth data of the channel is characterized by, In step (3), the value range of N is from 3 to the number of points-1, and point B needs to be removed from point set P.

5. The method for extracting the boundary of channel depth data as described in claim 1, characterized in that, In step (4), point C needs to be removed from point set P.

6. The method for extracting the boundary of channel depth data as described in claim 1, characterized in that, In step (5), point D needs to be removed from point set P, and point A is added to point set P.

7. The method for extracting the boundary of channel depth data as described in claim 1, characterized in that, In steps (5) and (6), after completing the angle sorting, the line with the maximum angle is tested for intersection with the existing boundary line. If it intersects, the candidate point forming the maximum angle is excluded.

Images