An electronic chart system water depth point intensive query method

Abstract

The present application relates to a kind of electronic chart system water depth point dense query method, can solve the defect of the limited availability of water depth point query function in current electronic chart system.It includes the following steps:(1) water depth data extraction, read water depth layer data from chart file, obtain water depth point record set and store.(2) gridding processing, utilize the scattered water depth data to establish gridding model, set step in horizontal, vertical two directions, select appropriate starting point to draw grid.(3) water depth data reconstruction, first, adopt the missing value completion technique to complete the missing value in water depth grid, form complete water depth grid model, then adopt super-resolution technology to improve water depth point density, form high-resolution water depth grid model.(4) water depth data query, when user clicks screen, obtain the water depth grid model obtained by reconstruction, compare the geographic coordinates of click place with the geographic coordinates in grid model, return query result.

Description

Technical Field

[0001] This invention relates to the field of marine information processing, and in particular to a method for dense querying of water depth points in an electronic nautical chart system. Background Technology

[0002] In electronic nautical charts, depth points are data points representing the depth of specific locations on the bottom of the ocean or body of water. This data is crucial for navigational safety, hydrological surveys, marine engineering, and marine scientific research. In electronic nautical chart systems, the depth value of a depth point is typically displayed at the screen coordinates corresponding to its location in the chart file. The depth point query function in electronic nautical chart systems involves clicking on the screen coordinates; if a depth point exists at that location, the query returns its value; otherwise, it returns a null value. However, because the number of depth points in electronic nautical chart data files is limited, depth values ​​are only available at a limited number of locations, and the query results for most locations are null, limiting the usability of the query function. Summary of the Invention

[0003] The purpose of this invention is to address the limited usability of depth point query functions in electronic nautical chart systems by proposing a dense depth point query method for these systems. This method is a complete dense depth point query scheme, including depth data extraction, depth data gridding, depth data reconstruction, and depth data query. The technical solution is as follows:

[0004] A method for dense querying of water depth points in an electronic nautical chart system:

[0005] (1) Water depth data extraction. Water depth data is read from the nautical chart file. The electronic nautical chart data is divided into layers. The water depth data is stored in the "SOUNDG" layer of the nautical chart, which is a point layer. The layer data is read to obtain the set of water depth point records. The data in the set of water depth points is extracted and stored.

[0006] (2) Grid processing. A grid model is established using scattered water depth data. A specified step size is set in both the horizontal and vertical directions. A suitable starting point is selected to draw straight lines to form a grid. Each water depth point is dispersed in different grids to generate a grid model.

[0007] (3) Depth Data Reconstruction. Depth data reconstruction is divided into two stages: missing value completion and super-resolution. First, missing value completion technology is used to complete the missing values ​​in the depth grid to form a complete depth grid model. Then, super-resolution technology is used to increase the density of depth points to form a high-resolution depth grid model. Missing value completion is performed using a conditional generative adversarial network, and super-resolution is performed using a Transformer network.

[0008] (4) Water depth data query. When the user clicks the screen, the screen coordinates of the query point are converted into projected coordinates and then into geographic coordinates according to the geographic coordinate system used by the electronic nautical chart system. The reconstructed water depth grid model is obtained, and the geographic coordinates of the clicked point are compared with the geographic coordinates in the grid model. If the comparison is successful, the geographic coordinates and water depth point values ​​at that point are used as the query results and highlighted on the nautical chart.

[0009] The beneficial effects of this invention are:

[0010] (1) This invention ensures the integrity of water depth grid data through spatial interpolation technology, improves the accuracy of water depth data through super-resolution technology, and finally realizes dense query of water depth points.

[0011] (2) The present invention can make more effective use of existing water depth data resources through dense query method, reduce the need for additional data collection, reduce data acquisition costs, and improve data processing efficiency.

[0012] (3) This invention allows users to quickly obtain water depth information through simple click operations, which improves the convenience and intuitiveness of user interaction. Attached Figure Description

[0013] Figure 1 This is an overall structural diagram of the present invention; Detailed Implementation

[0014] To make the technical solution of the present invention clearer, the present invention will be further described below with reference to the accompanying drawings. The present invention is implemented in specific steps:

[0015] (1) Water depth data extraction

[0016] Electronic nautical chart data is divided into layers. Water depth data is stored in the "SOUNDG" layer of the chart, which is a point layer. This data needs to be extracted for elevation model reconstruction. The water depth point extraction process is as follows: Read the "SOUNDG" layer of the electronic nautical chart data, which is the water depth point layer; obtain the layer record set; traverse the set to obtain the geometric information of each record object, i.e., the water depth point set; extract the water depth point set data and store the set data.

[0017] (2) Mesh processing

[0018] A gridded model was built using scattered water depth data. The grid structure is simple, with uniform resolution, facilitating retrieval and storage, and is easily integrated with nautical data, making it highly practical. The gridding method used is as follows: first, a predetermined step size is set in both the horizontal and vertical directions; then, straight lines are drawn from appropriate starting points to form a grid. Each water depth point is distributed across different grid cells, generating the grid model. The generated grid model is represented by the following matrix:

[0019]

[0020] Among them, Z 0,0 Z represents the elevation value of the upper left corner of the nautical chart. i,j This indicates the elevation value of the lower right corner of the nautical chart.

[0021] (3) Water depth data reconstruction

[0022] Water depth data reconstruction involves two stages: missing value completion and super-resolution. Missing value completion refers to calculating the elevation of an unknown point based on several adjacent known elevation data points. In real-world ocean elevation measurement scenarios, the measured elevation data is often scattered and cannot directly form a regular DEM, requiring missing value completion for reconstruction. Super-resolution refers to obtaining high-resolution water depth data from low-resolution water depth data. This method can obtain high-resolution water depth data without increasing additional measurement costs and ensuring reconstruction quality. Using deep learning methods to reconstruct water depth data, treating each coordinate point as a pixel position in a two-dimensional image and the height value as the image grayscale value, the water depth data reconstruction can be transformed into a two-dimensional image reconstruction.

[0023] Missing value completion is performed using a conditional generative adversarial network (GAN). The network is trained using a publicly available ocean elevation dataset. Original water depth images are cropped, and random sampling is performed proportionally to obtain sampled data, which serves as scattered water depth data. This sampled data is input as conditional data into the network's generator, guiding it to generate output that conforms to the conditional description. The sampled data, original data, and generated data are concatenated and fed into a discriminator. The discriminator needs to determine whether it is original or generated data, and also whether it conforms to the distribution of the sampled data. Classic models such as CycleGAN and StyleGAN can be used for the network model. To better recover terrain features, slope and aspect indicators from the DEM data are incorporated into the loss function, making the generated image more consistent with the original terrain. The z-axis is used to... i,j The elevation value at position (i,j) in the DEM data is represented by , cellsize represents the image resolution, dx and dy represent the horizontal and vertical slopes respectively, and slope is the slope value at position (i,j). i,j The formula is expressed as:

[0024]

[0025] Aspect represents the direction of the slope, typically expressed as the angle from the x-axis clockwise to the slope vector, ranging from 0 to 360°. Each DEM dataset can generate a slope map and an aspect map. The more similar the generated slope map and aspect map are to the actual data, the more the generated terrain matches the original terrain.

[0026] Super-resolution was achieved using a Transformer network. The network was trained using a publicly available ocean elevation dataset. Low-resolution depth data was obtained by downsampling the original water depth image and then fed into the Transformer network to guide the generation of a high-resolution image that matches the original image. The network employs a Swin Transformer-based reconstruction model, consisting of a shallow feature extraction module, a deep feature extraction module, and a reconstruction module. The shallow feature extraction module, composed of convolutional layers, extracts low-frequency information from the image. The deep feature extraction module, composed of Swin Transformer blocks, extracts deep information from the image. The reconstruction module fuses the shallow and deep features to generate the final high-resolution image. The Transformer overcomes the problem that convolution cannot focus on global terrain, improving the model's expressive power. Furthermore, to address the irreversible texture degradation caused by downsampling, a reference image mechanism is introduced. An image with the same terrain is selected as a reference image. Feature matching is performed between the low-resolution image and the reference image, providing additional terrain information for high-resolution image reconstruction and helping to improve the detail and clarity of the target image. In the improved attention module, the low-resolution image is upsampled to obtain an image of the same size as the high-resolution image, corresponding to Q in the attention mechanism. A reference image of the same size and from the same terrain region is selected, corresponding to V in the attention mechanism. The reference image is then downsampled and upsampled sequentially to obtain a reference image of the same size but with degraded features, corresponding to K in the attention mechanism.

[0027] (4) Water depth data query

[0028] The water depth data query function allows users to select and query water depth points on the nautical chart. The query process is as follows: After enabling the query function, the user clicks on the nautical chart interface to obtain the screen coordinates of the clicked point; according to the geographic coordinate system used by the electronic nautical chart system, the screen coordinates of the query point are converted into projected coordinates and then into geographic coordinates; the reconstructed water depth grid model is obtained, and the geographic coordinates of the clicked point are compared with the geographic coordinates in the grid model. If the comparison is successful, the geographic coordinates and water depth point value at that point are used as the query result and highlighted on the nautical chart.

Claims

1. A method for dense querying of water depth points in an electronic nautical chart system, characterized in that, Includes the following steps: (1) Water depth data processing: Read water depth data from the nautical chart file. The water depth data is stored in the "SOUNDG" layer of the nautical chart, which is a point layer. Read the layer, obtain the set of water depth point records, extract the data in the set of water depth points and store it. Use the scattered water depth data to build a grid model. Specifically, set a specified step size in the horizontal and vertical directions, select the starting point to draw a straight line to form a grid. Each water depth point is scattered in different grids to generate a grid model. (2) Water depth data reconstruction: Deep learning methods are used to fill in missing values ​​in the grid model and improve its resolution. Among them, missing value filling is performed using a conditional generative adversarial network. The original water depth image is randomly sampled to obtain sampled data as scattered water depth data. The sampled data is used as conditional data input into the generator of the network to guide the generator to generate output that meets the condition description. In terms of network model, the classic CycleGAN and StyleGAN models are adopted, and the slope and aspect indicators in the DEM data are introduced into the loss function. The super-resolution uses a reconstruction model based on Swin Transformer. The reconstruction model is divided into a shallow feature extraction module, a deep feature extraction module and a reconstruction module. The shallow feature extraction module consists of convolutional layers and introduces a reference image mechanism. Images with the same terrain are selected as reference images, and feature matching is performed between the low-resolution image and the reference image. (3) Water depth data query: Supports users to select and query water depth points on the electronic nautical chart interface.

2. The method for dense querying of water depth points in an electronic nautical chart system according to claim 1, characterized in that, In step (3), the specific process of water depth data query is as follows: water depth data query, the water depth data query function supports users to select and query water depth point values ​​on the nautical chart. The query process is as follows: after the query function is enabled, the user clicks on the nautical chart interface to obtain the screen coordinates of the clicked point. According to the geographic coordinate system used by the electronic nautical chart system, the screen coordinates of the query point are converted into projected coordinates and then into geographic coordinates. Obtain the reconstructed water depth grid model, compare the geographic coordinates of the clicked point with the geographic coordinates in the grid model, and if the comparison is successful, use the geographic coordinates and water depth value of that point as the query result and highlight them on the nautical chart.

Images