Data access method for realizing rapid display and browsing of massive point clouds

Abstract

The invention relates to the technical field of surveying and mapping information, and discloses a data access method for quickly displaying and browsing massive point clouds, which comprises the following steps: sequentially reading obtained original point cloud data files into a memory in batches; reconstructing the batch data in the memory into spatial rule data with an octree structure; and sequentially writing the reconstructed spatial rule data into a target file in batches to obtain a customized massive point cloud data file. And based on the customized massive point cloud data file, only the spatial data corresponding to the root node of the octree is retrieved and loaded during display, and only the spatial data in the corresponding range of the view area is retrieved and loaded during browsing. The data access method for quickly displaying and browsing the massive point clouds has the advantages of quickly displaying the point cloud data with ultra-large data volume on a common computer and realizing smooth browsing.

Description

Technical Field

[0001] The present invention relates to the field of surveying and mapping information technology, and particularly to a data access method for realizing fast display and browsing of massive point clouds. Background Art

[0002] Three-dimensional lidar technology is a non-contact active technology for rapidly acquiring a three-dimensional dense point cloud on the surface of an object. It can directly acquire massive, irregularly spatially distributed three-dimensional point clouds with three-dimensional coordinates and certain attributes. With the continuous development of technology, it has become an important new means for acquiring data in fields such as digital cities and traditional surveying and mapping, and plays an increasingly important role in fields such as global change, smart cities, resource surveys, environmental monitoring, and basic surveying and mapping. The application of three-dimensional laser point cloud data is also becoming more and more extensive.

[0003] Among them, three-dimensional laser point cloud data mainly includes three-dimensional coordinates (X, Y, Z) and certain attributes (reflection intensity, color information, echo information, etc.). A point cloud data file with the above information data often has an extremely large data volume, and it is not uncommon for it to be in the order of dozens or hundreds of GB. On the other hand, as the most basic function of the application of three-dimensional laser point cloud data, there are also burdens in the display and browsing of extremely large data volumes. General software has the following problems in the application of massive point cloud data: (1) It can only load and display smaller data, and cannot display or browse massive point cloud data of several G or hundreds of G. (2) It depends on a super-powerful computer to barely display, but the browsing is not smooth or even stuck. (3) Split the original point cloud data file into numerous small data files (usually hundreds or thousands of small files) for display, and the access time is extremely long, and the visual effect is poor.

[0004] If the above existing problems can be solved, it will effectively promote the application of massive point cloud data and have great economic and social value; Therefore, in view of the above existing problems, a data access method for realizing fast display and browsing of massive point clouds is proposed. Summary of the Invention

[0005] (I) Technical Problems to be Solved In view of the deficiencies of the prior art, the present invention provides a data access method for realizing fast display and browsing of massive point clouds, which has the advantage of enabling point cloud data with an extremely large data volume to be quickly displayed on an ordinary computer and achieving smooth browsing.

[0006] (II) Technical Solutions The technical solution of the present invention for solving the above technical problems is as follows: A data access method for realizing fast display and browsing of massive point clouds, comprising: Batch read the obtained original point cloud data files into memory in sequence; Reconstruct the batch point cloud data in memory into spatial regular data with an octree structure; Batch write the reconstructed spatial regular data into the target file in sequence, and thus obtain a custom massive point cloud data file.

[0007] Furthermore, the data reconstruction process is to reconstruct the spatial regular data with an octree structure batch by batch in sequence, and the writing process of the spatial regular data is to write into the target file batch by batch in sequence.

[0008] Furthermore, the custom massive point cloud data file is a single file.

[0009] Furthermore, the obtained custom massive point cloud data file includes a data standard header, a data retrieval header, and a data area; Among them, the data standard header is used to record file description information, the data retrieval header is used to record data retrieval information, and the data area is used to record original point cloud data information.

[0010] Furthermore, the data retrieval information includes octree space information and retrieval information.

[0011] Furthermore, the original point cloud data information includes the coordinates X, Y, Z of the original point cloud and related attribute information.

[0012] Furthermore, after batch writing the reconstructed spatial regular data into the target file in sequence to obtain the custom massive point cloud data file, it further includes: scheduling the data in the data area according to the data retrieval header of the obtained custom massive point cloud data file for display.

[0013] Furthermore, after reading the data retrieval header data of the custom massive point cloud data file, it further includes: Parsing the data retrieval header to obtain root node index information; Reading the point cloud data corresponding to the index into memory; Loading the in-memory point cloud data to the interface for display.

[0014] Furthermore, after reading the data retrieval header data of the custom massive point cloud data file, it further includes: Calculating the size of the projected pixel area of each index data block on the screen; Selecting the data blocks whose projected pixel area meets the adjustment; Reading the point cloud data corresponding to the index into memory; Loading the obtained data and releasing the non-current data.

[0015] The beneficial effects of the present invention are: The present invention converts an ultra-large original point cloud data file into a custom massive point cloud data file. This file organizes data in a spatial index structure, featuring efficient access and storage. Moreover, the conversion process directly writes to a single result file without generating multiple intermediate result temporary files. The custom massive point cloud data file is a single large file, enabling a very fast conversion speed on ordinary computers. Additionally, the operations of loading, displaying, or browsing the custom massive point cloud data file are also highly efficient, achieving the effects of instantly displaying the ultra-large point cloud file and smooth browsing without lag. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flowchart for generating a custom massive point cloud data file according to the present invention; Figure 2 It is a flowchart for the display and browsing method of the massive point cloud data according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] Next, in combination with the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0018] From Figure 1 and Figure 2 A data access method for realizing fast display and browsing of massive point clouds is given, including: Sequentially batch read the obtained original point cloud data file into the memory; Reconstruct the batch point cloud data in the memory into spatially regular data with an octree structure; The above process is because the point cloud data recorded in the original point cloud data file is spatially irregular. Therefore, the irregular point cloud data is organized into spatially regular data with a spatial index. For point cloud data in three-dimensional space, the octree structure is more applicable; Among them, the octree structure continuously divides the space into sub-spaces: equally divides the space into eight sub-spaces as child nodes in the three dimensions of X, Y, and Z, and continues to divide the child nodes in this way until the limiting condition is met and the division stops, thus obtaining spatially regular data with an octree structure; Sequentially batch write the reconstructed spatially regular data into the target file, that is, a custom massive point cloud data file is obtained, and the custom massive point cloud data file is a single file.

[0019] Further, the data reconstruction process is to reconstruct the spatially regular data with an octree structure batch by batch in sequence, and the writing process of the spatially regular data is to write to the target file batch by batch in sequence; Since the original point cloud data file may have an extremely large data volume, several GB, dozens of GB or even larger, the capabilities of ordinary computers are not sufficient to read all the data into the memory at one time. Therefore, a strategy of reading file data in batches is adopted. The data of each batch is reconstructed according to the octree method, organized into spatially regular and indexed data, and then written to the result file batch by batch until the end. This process does not generate temporary files of multiple intermediate results, so the speed is very fast.

[0020] Further, the obtained custom massive point cloud data file includes a data standard header, a data retrieval header, and a data area; Among them, the data standard header is used to record file description information; The data retrieval header is used to record data retrieval information. The data retrieval information includes octree space information and retrieval information, which are used for rapid positioning and searching for quick data access; The data area is used to record the original point cloud data information. The original point cloud data information includes the coordinates X, Y, Z of the original point cloud and related attribute information, etc.; The above data structure has the characteristics of fast and convenient data access, and can help to quickly display and browse the point cloud.

[0021] Further, after the reconstructed spatially regular data is written to the target file batch by batch in sequence to obtain a custom massive point cloud data file, it further includes: scheduling the data in the data area according to the data retrieval header of the obtained custom massive point cloud data file for display; Preferably, it further includes: scheduling and loading the data corresponding to the root node for display according to the data retrieval header of the obtained custom massive point cloud data file. Compared with the entire data file, it is extremely small, so the data display effect is extremely fast.

[0022] Further, after reading the data retrieval header data of the custom massive point cloud data file, it further includes: Parsing the data retrieval header to obtain the root node index information; Reading the point cloud data corresponding to the index into the memory; Loading the in-memory point cloud data to the interface for display; The above operations are display operations. Since the interval data corresponding to the root node is the downsampled data within the entire range, only reading the root node data for display solves the problem that the large amount of data cannot be accessed in the memory, and the overall data can also be visually seen; Further, after reading the data retrieval header data of the custom massive point cloud data file, it further includes: Calculate the size of the projected pixel area of each index data block on the screen; Select the data blocks whose projected pixel areas meet the adjustment; Read the point cloud data corresponding to the index into the memory; Load the acquired data and release the non-current data; The above operations are browsing operations, and the specific operations are as follows: Select the data blocks to be loaded according to the size of the projected pixel areas of each index data block on the screen, then locate the position of the data through the retrieval header part of the custom massive point cloud data file, and then read the corresponding part of the point cloud data and load it into the view window for display, rather than traversing all the data in the file. Therefore, the browsing speed is very fast; Moreover, when browsing to other data ranges, such as zooming in, moving, rotating, etc., similar to the above method, load only the point cloud data that meets the conditions within the current view window range, unload the non-current data from the memory to ensure that the current memory data is appropriate and not overloaded. Its data volume is small, so it avoids loading too large a data volume; at the same time, release the memory point cloud data that is no longer displayed. Through the above method, it can ensure that the limited memory will not be overloaded, and dynamically only access and store an appropriate number of point cloud data, solving the problem of lag when browsing ultra-large data.

[0023] Specifically, as can be seen from the above technical solution, the present invention converts an ultra-large original point cloud data file into a custom massive point cloud data file, which organizes data in a spatial index structure and has the characteristics of efficient access. And the conversion process is to directly write to a single result file, which will not generate multiple intermediate result temporary files. Moreover, the custom massive point cloud data file is a single large file, making its conversion speed very fast on a common computer, and the operations of loading, displaying or browsing the custom massive point cloud data file are also very efficient, achieving the effect of quickly opening and displaying an ultra-large point cloud file and smooth browsing without lag.

[0024] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article or device including the element.

[0025] Although embodiments of the present invention have been shown and described, those of ordinary skill in the art will appreciate that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A data access method for realizing fast display and browsing of massive point clouds, characterized in that: Including: Batch read the obtained original point cloud data files into memory in sequence; Reconstruct the batch point cloud data in memory into spatially regular data with an octree structure; Batch write the reconstructed spatially regular data into the target file in sequence, thus obtaining a custom massive point cloud data file.

2. The data access method for realizing fast display and browsing of massive point clouds according to claim 1, characterized in that: The data reconstruction process is to reconstruct the spatially regular data with an octree structure batch by batch in sequence, and the writing process of the spatially regular data is to write into the target file batch by batch in sequence.

3. A data access method for realizing rapid display and browsing of massive point clouds according to claim 1, characterized in that: The custom massive point cloud data file is a single file.

4. A data access method for realizing rapid display and browsing of massive point clouds according to any one of claims 1-3, characterized in that: The specific structure of the obtained custom massive point cloud data file includes a data standard header, a data retrieval header, and a data area; Among them, the data standard header is used to record file description information, the data retrieval header is used to record data retrieval information, and the data area is used to record original point cloud data information.

5. A data access method for realizing rapid display and browsing of massive point clouds according to claim 4, characterized in that: The data retrieval information includes octree space information and retrieval information.

6. A data access method for realizing fast display and browsing of massive point clouds according to claim 4, characterized in that: The original point cloud data information includes the coordinates X, Y, Z of the original point cloud and related attribute information.

7. A data access method for realizing rapid display and browsing of massive point clouds according to claim 4, characterized in that: After batch writing the reconstructed spatially regular data into the target file in sequence to obtain the custom massive point cloud data file, it further includes: scheduling the data in the data area according to the data retrieval header of the obtained custom massive point cloud data file for display.

8. A data access method for realizing rapid display and browsing of massive point clouds according to claim 7, characterized in that: After reading the data retrieval header data of the custom massive point cloud data file, it further includes: Analyze the data retrieval header to obtain root node index information; Read the point cloud data corresponding to the index into memory; Load the in-memory point cloud data to the interface for display.

9. A data access method for realizing rapid display and browsing of massive point clouds according to claim 7, characterized in that: After reading the data retrieval header data of the custom massive point cloud data file, it further includes: Calculate the size of the projected pixel area of each index data block on the screen; Select the data blocks whose projected pixel areas meet the adjustment; Read the point cloud data corresponding to the index into memory; Load the obtained data and release the non-current data.

Images