WO2020154975A1 - 用于点云数据的存储方法、回放方法及计算机可读介质 - Google Patents
用于点云数据的存储方法、回放方法及计算机可读介质 Download PDFInfo
- Publication number
- WO2020154975A1 WO2020154975A1 PCT/CN2019/073984 CN2019073984W WO2020154975A1 WO 2020154975 A1 WO2020154975 A1 WO 2020154975A1 CN 2019073984 W CN2019073984 W CN 2019073984W WO 2020154975 A1 WO2020154975 A1 WO 2020154975A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- storage method
- data packet
- sensor
- information
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/51—Indexing; Data structures therefor; Storage structures
Definitions
- the present invention relates to the field of information storage technology, and in particular to a storage method, playback method and computer readable medium for point cloud data.
- the laser ranging system obtains three-dimensional information of the scene through the sensing sensor.
- the basic principle is to actively emit laser pulse signals to the detected object and obtain the reflected pulse signals, and calculate the depth information of the distance detector of the detected object according to the time difference between the transmitted signal and the received signal; based on the laser ranging system Obtain the angle information of the measured object relative to the laser ranging system; combine the aforementioned depth and angle information to obtain a large number of detection points (called point clouds), based on the point cloud, the relative laser ranging of the measured object can be reconstructed The spatial three-dimensional information of the system.
- point clouds detection points
- Point cloud data has many applications. After the application layer program obtains the point cloud from the network layer, it needs to be dumped in the form of a file for subsequent algorithm related personnel to conduct research and development, or perform data through specific software. Replay.
- one aspect of the embodiments of the present invention provides a storage method for point cloud data.
- the method uses at least a three-tier structure for storage.
- the three-tier structure includes a file layer and a data frame.
- Layer and packet layer wherein the method includes:
- the file layer including a file header, a device information table and at least one of the data frames;
- each frame of data Acquiring each frame of data and storing it in the data frame layer, the data frame layer containing information of each of the data frames in the file layer, wherein each of the data frames includes a frame header and at least one data packet,
- the frame header contains address information of the data frame
- the data packet layer contains the information of each data packet in the data frame layer, wherein each data packet includes a data packet header and a dot Cloud data, the data packet header contains the device information and time stamp information of the sensor to which the current data packet belongs.
- Another aspect of the embodiments of the present invention provides a point cloud file playback method for playing back the point cloud file stored according to the foregoing storage method, and the playback method includes: according to the address information contained in the frame header Positioning to a specific data frame for playback, or positioning to a specific data packet for playback according to the time stamp information contained in the data packet header.
- Another aspect of the present invention provides a computer-readable medium having a computer program stored on the computer-readable medium, and the computer program executes the point cloud data storage method of the embodiment of the present invention when the computer program is running.
- Another aspect of the present invention provides a computer-readable medium having a computer program stored on the computer-readable medium, and the computer program executes the point cloud file playback method of the embodiment of the present invention when the computer program is running.
- the point cloud data storage method, playback method, and computer readable medium of the present invention store point cloud data in at least a three-layer structure including a file layer, a data frame layer, and a data packet layer, allowing users to instantly play the point cloud. It can be stored online, and the stored point cloud files have strong robustness. Even if the application layer program crashes during operation, the remaining point cloud files can still be played; and the point cloud files stored according to this storage method only need to be very robust. It can be played back in a short analysis time.
- Fig. 1 shows a flowchart of a storage method for point cloud data according to an embodiment of the present invention
- Fig. 2 shows a schematic diagram of a hierarchical structure of a point cloud file stored in a storage method for point cloud data according to an embodiment of the present invention.
- one aspect of the present invention provides a storage method for point cloud data, which adopts at least a three-tier structure for storage.
- the three-tier structure includes a file layer and a data frame.
- the storage method includes: obtaining a data frame and storing it in the file layer, the file layer including a file header, a device information table, and at least one of the data frame; obtaining each frame of data and storing it in the file layer
- the data frame layer, the data frame layer includes information of each of the data frames in the file layer, wherein each of the data frames includes a frame header and at least one data packet, and the frame header includes The address information of the data frame; the data packet in the data frame is acquired and stored in the data packet layer, and the data packet layer contains the information of each data packet in the data frame layer, wherein each of the data
- the packet includes a data packet header and point cloud data, and the data packet header contains device information and time stamp information of the sensor to which the current data packet belongs.
- another aspect of the present invention provides a point cloud file playback method for playing back the point cloud file stored according to the above storage method, and the playback method includes: according to the frame header contained in the The address information is located to a specific data frame for playback, or, according to the time stamp information contained in the data packet header, a specific data packet is located for playback.
- the point cloud data may come from at least one radar in a distributed radar system.
- the distributed radar system includes a host computer and multiple radars.
- the multiple radars are distributed at different positions on the movable platform to detect object information in different positions/directions.
- the host computer integrates the information of objects detected by multiple radars. Processing to understand the object information in the surrounding environment. Since each lidar has a certain field of view (FOV), and each lidar has a certain emission direction angle, the field of view angles of adjacent radars partially overlap or do not overlap, passing through the field of view of multiple lidars Splicing can detect the surroundings of the movable platform, so as to understand the environmental information around the movable platform.
- FOV field of view
- the distributed radar system may also include a center board that receives instructions sent by the upper computer and distributes them to multiple radars; or, the center board may also receive data packets sent by the radar and send them to the upper computer ; Further, the center board parses the data packets from each lidar, aligns the data obtained by analyzing each data packet, and then sends the aligned data. The center board aligns the timing of each data packet, including obtaining a synchronization timing signal input from the timing port of the center board, which also provides timing for each lidar. The center board analyzes the received data packets from each lidar, obtains the time stamp data of laser emission and echo, and aligns the time stamp data in time sequence based on the synchronization timing signal.
- the center board can also stitch data from various lidars. In this way, the data from each lidar will be converted into a data packet, thereby reducing the amount of calculation on the host computer.
- the splicing of the center board to the host computer can be performed according to preset port rules, which correspond to the overlap of the field of view of the lidar and the position information of the lidar.
- the upper computer may include one or more processors for receiving data sent by multiple radars, processing the data, and issuing instructions to control the work of the radars and other modules.
- the upper computer has multiple radar interfaces, and the radar can be connected to the radar interface through a transmission cable to connect the radar to the upper computer so that the upper computer can receive radar data and control the radar.
- the point cloud data storage method, playback method, and computer readable medium of the embodiments of the present invention store point cloud data in at least a three-layer structure including a file layer, a data frame layer, and a data packet layer. It can be stored online in the cloud, and the stored point cloud files have strong robustness. Even if the application layer program crashes during operation, the remaining point cloud files can still be played; and the point cloud files stored according to the storage method It only takes a short analysis time to play back.
- Fig. 1 shows a flowchart of a storage method 100 for point cloud data according to an embodiment of the present invention.
- the storage method 100 may be used to transfer point cloud data obtained from the network layer in the form of a file. For subsequent algorithm related personnel to conduct research and development, or for data playback through specific software.
- the storage method adopts at least a three-layer structure for storage, and the three-layer structure includes a file layer, a data frame layer, and a data packet layer.
- the format of the point cloud file stored according to the storage method 100 is named as the .lvx file.
- the hierarchical structure of the Lvx file is shown in FIG. 2.
- the Lvx file includes a file layer 201, a data frame layer 202, and data Cladding 203.
- the storage method 100 for point cloud data includes the following steps:
- a data frame is obtained and stored in the file layer 201.
- the file layer 201 includes a file header, a device information table, and at least one data frame.
- a file header, a device information table, and at least one data frame are stored in the file layer 201.
- the point cloud data is organized in the form of data frames to facilitate playback.
- all data frames can be stored at a frequency of 20 Hz, that is, the actual duration of a single frame is 50 ms.
- At least the file flag, protocol version information and check digit are obtained and stored in the file layer 201 as the file header.
- the data name, offset, and data type contained in the file header are shown in Table 1:
- the magic code is the file identifier, used to distinguish the file type, and also used to identify the start code of the file.
- Its data type can be a signed integer (int8_t), occupying 16 bytes .
- the version (version) represents the protocol version information
- the data type can be an unsigned integer (uint8_t)
- crc32 represents the checksum of the file header, which is used to check the correctness of the data. Of course, in addition to this, other types of check methods can also be used.
- the sensor information to which the point cloud data belongs is acquired and stored in the device information table. Further, the number of sensors to which the point cloud data belongs and the device information of each sensor are acquired to be stored in the device information table. According to the information in the device information table, the user can obtain the device information of the sensor used when the point cloud file is recorded.
- Table 2 Exemplarily, the data name, offset, and data type included in the device information table are shown in Table 2:
- info count represents the number of sensors to which the point cloud data in the point cloud file belongs.
- the sensor that generates point cloud data is used to sense external environmental information, such as distance information, azimuth information, reflection intensity information, speed information, etc. of environmental targets. It can be a lidar, or other radars or ranging devices.
- the number of the sensors can be multiple, and the multiple sensors can be distributed at different positions of the movable platform to detect object information in different positions/directions around the movable platform.
- the device information of each sensor is stored in the Device info list (device information list).
- the device information list includes the information of N sensors, and the device information of each sensor is shown in Table 3:
- the sensor serial number When the sensor is a lidar, the sensor serial number includes a radar serial number (lidar sn code), and specifically includes a radar broadcast code (LiDAR Broadcast Code).
- the sensor serial number can be used to indicate the sensor to which the point cloud data belongs, that is, the sensor that generates the point cloud data.
- the hub sn code is the serial number of the hub connected to each sensor.
- the central board includes the central control circuit data summary, which plays the role of data summary and forwarding. It can forward the information from the upper computer to each sensor, and it can also send the information from the sensor to the upper computer.
- the central board has time The alignment function can align the data of multiple sensors in time. However, it should be noted that the sensor can directly exchange information with the upper computer, or exchange information with the upper computer through the center board. Exemplarily, if it is not detected that the sensor is connected to the center board, the serial number of the center board is set to empty. In other embodiments, for a radar system that has been determined not to include a center board, the device information table may not include information about the center board.
- the device index is the index of the current sensor in the device information list. It corresponds to the index in the data packet header, and can be used to distinguish the sensor to which each data packet belongs, so that the user can classify the point cloud for each sensor.
- Device type contains sensor device type information.
- the device type information includes sensor type or center plate type. Exemplarily, 0 represents the center plate, and 1, 2, and 3 respectively represent three different radar types.
- the angle information of the sensor includes the roll, pitch, and yaw of the sensor; the position information of the sensor includes the x, y, and z axis coordinates of the sensor.
- the information type of the angle and position information of the sensor may be a floating point type (float).
- each frame of data is acquired and stored in the data frame layer 202.
- the data frame layer 202 includes information of each data frame in the file layer 201.
- Frame Header some necessary information of the current data frame is obtained and stored as a frame header (Frame Header).
- the information stored in each frame header is shown in Table 4:
- current offset is the start offset address of the current frame
- next offset is the start offset address of the next frame. Since the frame header contains the offset address of the current frame and the offset address of the next frame, the structure of all data frames is similar to a linked list, and users can learn the information of the current frame and the next frame from the frame header of the current frame. Information in order to quickly traverse the entire file.
- Frame index is the index of the current frame. Specifically, the index of the current frame is acquired to be stored in the frame header. Using the frame index can be used to quickly access a specific data frame.
- Pack cnt packet count, number of data packets
- the storage frequency of the data frame is 20 Hz, that is, the actual time corresponding to the data in each frame is 50 ms.
- the number of data packets in each frame may not be equal, and there may be certain fluctuations.
- a data packet in a data frame is acquired and stored in the data packet layer 203, and the data packet layer 203 contains the information of each data packet in the data frame layer 202, wherein each The data packet includes a data packet header and point cloud data, and the data packet header contains device information and time stamp information of the sensor to which the current data packet belongs.
- the data packet layer contains a large amount of point cloud data, and users can use the data in the data packet to apply self-developed algorithms to complete more operations.
- a device index (device index) is obtained to be stored in the data packet header.
- the device index is mapped to the index in the device information list one by one, and is used to distinguish the sensor to which the current data packet belongs. That is, multiple data packets in each data frame can belong to different sensors, but the point cloud data in each data packet comes from the same sensor.
- the method further includes obtaining protocol version information (version) to be stored in the data packet header.
- the protocol version is a data packet protocol version (Package Protocol Version).
- the method further includes acquiring the port address and the sensor ID connected to the sensor to store in the data packet header.
- the port address is the slot id (port id) shown in Table 5, which represents the address of the port used to connect the sensor.
- the slot id is 0 by default. If the sensor is connected to the central board and the central board is connected to the upper computer, the slot id is the central board port number.
- the sensor ID is a radar ID (LiDAR id).
- the serial number of the single radar is 0, and the serial number of the three-in-one radar is 0, 1, and 2, respectively.
- RSVD Reserved
- Status code is used to indicate the status of the sensor when it generates point cloud data.
- the method further includes obtaining a timestamp type (timestamp type) and a timestamp (timestamp) to be stored in the data packet header.
- a time stamp is stored in the data packet to indicate the time of the first point in the data packet. The user can locate a specific data packet for playback according to the timestamp.
- the time stamp type represents the type of time stamp. Exemplarily, the unit of the time stamp is nanoseconds (Nanosecond), or the format of the time stamp is Coordinated Universal Time (UTC Format).
- the method further includes obtaining data type (data type) information to be stored in the data packet header.
- the data type information is used to indicate the coordinate format of the original point cloud data.
- the data type information includes a Cartesian Coordinate type or a Spherical Coordinate type. For example, you can use 0 to represent the Cartesian coordinate type and 1 to represent the spherical coordinate type.
- each data packet contains 100 points of point cloud data. Therefore, correspondingly, in the storage method 100 according to the embodiment of the present invention, 100 points of point cloud data are stored in each of the data packets, and the point cloud data of each point is shown in Table 6:
- the point cloud data of each point includes the coordinates and intensity of each point.
- the coordinates of each point are the X-axis, Y-axis and Z-axis coordinates calculated by external parameters, and the data type is floating point.
- the brightness value can represent the reflectivity of the object to be measured at this point, and its data type can be an unsigned integer. Based on a large amount of point cloud data, the spatial three-dimensional information of the measured object relative to the sensor can be reconstructed.
- the point cloud data storage method allows users to store the point cloud online while playing the point cloud.
- the stored point cloud file has strong robustness, even if the application layer program is running If it crashes, the remaining point cloud files can still be played; and the point cloud files stored according to this storage method can be played back in a short analysis time.
- the computer-readable medium may include, for example, a memory card of a smart phone, a storage component of a tablet computer, a hard disk of a personal computer, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a portable compact disk only Readable memory (CD-ROM), USB memory, or any combination of the above computer-readable media.
- the computer-readable medium may be any combination of one or more computer-readable media.
- the frame header contains the address information of the data frame, specifically including the start offset address of the current frame, the start offset address of the next frame and the frame index.
- the user passes the frame header information of each frame. To traverse the point cloud file, you can get the total number of frames and the offset address of each frame in the file.
- the user can locate a specific data frame for playback according to the address information contained in the frame header.
- each data packet contains a time stamp
- the user can locate a specific data packet for playback according to the time stamp information contained in the data packet header.
- the specific structure of the point cloud file can be referred to above, for the sake of brevity, it will not be repeated here.
- the point cloud file playback method can quickly playback the point cloud file, allowing the user to analyze the file and restore it to the original point cloud data and perform algorithm analysis.
- a computer-readable medium on which program instructions are stored, and are used to execute the embodiments of the present invention when the program instructions are run by a computer or a processor.
- the corresponding steps of the playback method for point cloud files may include, for example, a memory card of a smart phone, a storage component of a tablet computer, a hard disk of a personal computer, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a portable compact disk only Readable memory (CD-ROM), USB memory, or any combination of the above computer-readable media.
- the computer-readable medium may be any combination of one or more computer-readable media.
- the disclosed device and method may be implemented in other ways.
- the device embodiments described above are merely illustrative.
- the division of the units is only a logical function division, and there may be other divisions in actual implementation, for example, multiple units or components can be combined or It can be integrated into another device, or some features can be ignored or not implemented.
- the various component embodiments of the present invention may be implemented by hardware, or by software modules running on one or more processors, or by a combination of them.
- a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all of the functions of some modules according to the embodiments of the present invention.
- DSP digital signal processor
- the present invention can also be implemented as a device program (for example, a computer program and a computer program product) for executing part or all of the methods described herein.
- Such a program for realizing the present invention may be stored on a computer-readable medium, or may have the form of one or more signals. Such signals can be downloaded from Internet websites, or provided on carrier signals, or provided in any other form.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Television Signal Processing For Recording (AREA)
Abstract
一种用于点云数据的存储方法、回放方法及计算机可读介质,所述存储方法采用至少三层结构的方式进行存储,所述三层结构包括文件层、数据帧层和数据包层,其中所述文件层包括文件头、设备信息表和至少一个所述数据帧;所述数据帧层中,每个数据帧包括帧头和至少一个数据包,所述帧头中包含有数据帧的地址信息;所述数据包层中,每个数据包包括数据包头和点云数据,所述数据包头中包含有当前数据包所隶属的传感器的设备信息和时间戳信息。所述用于点云数据的存储方法允许用户实时存储点云数据,存储的点云文件鲁棒性强并且便于回放。
Description
本发明涉及信息存储技术领域,具体而言涉及一种用于点云数据的存储方法、回放方法及计算机可读介质。
激光测距系统通过感知传感器获得场景的三维信息。其基本原理为主动对被探测对象发射激光脉冲信号,并获得其反射回来的脉冲信号,根据发射信号和接收信号之间的时间差计算被测对象的距离探测器的深度信息;基于激光测距系统的已知发射方向,获得被测对象相对激光测距系统的角度信息;结合前述深度和角度信息得到海量的探测点(称为点云),基于点云即可以重建被测对象相对激光测距系统的空间三维信息。
点云数据有多方面的应用,应用层程序将点云从网络层获取到之后,需要将点云以文件的形式进行转存,以供后续的算法相关人员进行研发,或者通过特定软件进行数据回放。
然而,由于激光雷达技术的不成熟,目前还没有针对点云数据的公认的储存方案,使用者均将点云按照各自组内的需求甚至个人的使用习惯来进行储存。因此,点云文件的适用范围会变得十分狭窄,只能在组内甚至于个人之间使用,无法实现点云数据的交流。
发明内容
在发明内容部分中引入了一系列简化形式的概念,这将在具体实施方式部分中进一步详细说明。本发明的发明内容部分并不意味着要试图限定出所要求保护的技术方案的关键特征和必要技术特征,更不意味着试图确定所要求保护的技术方案的保护范围。
针对现有技术的不足,本发明实施例一方面提供了一种用于点云数据的存储方法,所述方法采用至少三层结构的方式进行存储,所述 三层结构包括文件层、数据帧层和数据包层,其中所述方法包括:
获取数据帧,存储于所述文件层,所述文件层包括文件头、设备信息表和至少一个所述数据帧;
获取每帧数据,存储于所述数据帧层,所述数据帧层包含所述文件层中每个所述数据帧的信息,其中,每个所述数据帧包括帧头和至少一个数据包,所述帧头中包含有数据帧的地址信息;
获取数据帧中的数据包,存储于所述数据包层,所述数据包层包含所述数据帧层中每个所述数据包的信息,其中,每个所述数据包包括数据包头和点云数据,所述数据包头中包含有当前数据包所隶属的传感器的设备信息和时间戳信息。
本发明实施例另一方面提供了一种点云文件的回放方法,用于回放根据上述存储方法所存储的点云文件,所述回放方法包括:根据所述帧头中包含的所述地址信息定位到特定的数据帧进行回放,或者,根据所述数据包头中包含的时间戳信息定位到特定的数据包进行回放。
本发明再一方面提供了一种计算机可读介质,所述计算机可读介质上存储有计算机程序,所述计算机程序在运行时执行本发明实施例的用于点云数据的存储方法。
本发明再一方面提供了一种计算机可读介质,所述计算机可读介质上存储有计算机程序,所述计算机程序在运行时执行本发明实施例的点云文件的回放方法。
本发明的点云数据的存储方法、回放方法及计算机可读介质,以包括文件层、数据帧层和数据包层的至少三层结构的方式存储点云数据,允许用户在播放点云时即可进行在线储存,存储的点云文件具有较强的鲁棒性,即使应用层程序在运行中崩溃,残存的点云文件仍然能够播放;并且,根据该存储方法存储的点云文件只需很短的解析时间即可进行回放。
本发明的下列附图在此作为本发明的一部分用于理解本发明。附图中示出了本发明的实施例及其描述,用来解释本发明的原理。
附图中:
图1示出了根据本发明一实施例的用于点云数据的存储方法的流程图;
图2示出了根据本发明一实施例的用于点云数据的存储方法所存储的点云文件的层级结构示意图。
为了使得本发明的目的、技术方案和优点更为明显,下面将参照附图详细描述根据本发明的示例实施例。显然,所描述的实施例仅仅是本发明的一部分实施例,而不是本发明的全部实施例,应理解,本发明不受这里描述的示例实施例的限制。基于本发明中描述的本发明实施例,本领域技术人员在没有付出创造性劳动的情况下所得到的所有其它实施例都应落入本发明的保护范围之内。
在下文的描述中,给出了大量具体的细节以便提供对本发明更为彻底的理解。然而,对于本领域技术人员而言显而易见的是,本发明可以无需一个或多个这些细节而得以实施。在其他的例子中,为了避免与本发明发生混淆,对于本领域公知的一些技术特征未进行描述。
应当理解的是,本发明能够以不同形式实施,而不应当解释为局限于这里提出的实施例。相反地,提供这些实施例将使公开彻底和完全,并且将本发明的范围完全地传递给本领域技术人员。
在此使用的术语的目的仅在于描述具体实施例并且不作为本发明的限制。在此使用时,单数形式的“一”、“一个”和“所述/该”也意图包括复数形式,除非上下文清楚指出另外的方式。还应明白术语“组成”和/或“包括”,当在该说明书中使用时,确定所述特征、整数、步骤、操作、元件和/或部件的存在,但不排除一个或更多其它的特征、整数、步骤、操作、元件、部件和/或组的存在或添加。在此使用时,术语“和/或”包括相关所列项目的任何及所有组合。
为了彻底理解本发明,将在下列的描述中提出详细的步骤以及详细的结构,以便阐释本发明提出的技术方案。本发明的较佳实施例详细描述如下,然而除了这些详细描述外,本发明还可以具有其他实施方式。
为了解决现有技术中存在的上述技术问题,本发明一方面提供一种用于点云数据的存储方法,其采用至少三层结构的方式进行存储,所述三层结构包括文件层、数据帧层和数据包层,其中所述存储方法包括:获取数据帧,存储于所述文件层,所述文件层包括文件头、设备信息表和至少一个所述数据帧;获取每帧数据,存储于所述数据帧层,所述数据帧层包含所述文件层中每个所述数据帧的信息,其中,每个所述数据帧包括帧头和至少一个数据包,所述帧头中包含有数据帧的地址信息;获取数据帧中的数据包,存储于所述数据包层,所述数据包层包含所述数据帧层中每个所述数据包的信息,其中,每个所述数据包包括数据包头和点云数据,所述数据包头中包含有当前数据包所隶属的传感器的设备信息和时间戳信息。
与此相对的,本发明另一方面提供一种点云文件的回放方法,用于回放根据上述存储方法所存储的点云文件,所述回放方法包括:根据所述帧头中包含的所述地址信息定位到特定的数据帧进行回放,或者,根据所述数据包头中包含的时间戳信息定位到特定的数据包进行回放。
示例性地,本发明实施例所提供的点云数据的存储方法中,所述点云数据可来自于分布式雷达系统中的至少一个雷达。所述分布式雷达系统包括上位机和多个雷达,多个雷达分布在可移动平台的不同的位置,用于检测不同位置/方向的物体信息,上位机根据多个雷达所检测物体信息进行综合处理,从而了解周围环境的物体信息。由于每个激光雷达都具有一定的视场角(FOV),并且每个激光雷达具有一定的发射方向角,相邻雷达的视场角部分重叠或者恰好不重叠,通过多个激光雷达的视场拼接,可以探测可移动平台周围,从而了解可移动平台周围的环境信息。
所述分布式雷达系统还可以包括中心板,所述中心板接收上位机发送的指令,并分发给多个雷达;或者,所述中心板还可以接收雷达发送的数据包,并发送给上位机;进一步地,中心板对来自各个激光雷达的数据包进行解析,对各个数据包解析获得的数据进行时序对齐,然后将对齐后的数据发送。中心板对各个数据包的时序对齐,包括,获得来自中心板时序端口输入的同步时序信号,该同步时序信号 也为各个激光雷达提供时序。中心板对接收到的,来自各激光雷达的数据包进行解析,获得激光发射和回波的时间戳数据,并基于同步时序信号将各时间戳数据在时序上对齐。在进行时间戳对齐的同时,中心板还可以对来自各个激光雷达的数据进行拼接。这样,来自各个激光雷达的数据会被整个成为一个数据包,以此减少上位机的运算量。中心板对上位机的拼接可以根据预设端口规则进行,预设端口规则与激光雷达的视场角交叠以及激光雷达的位置信息相对应。
上位机可以包括一个或多个处理器,用于接收多个雷达发送的数据,并对数据进行处理,以及下发指令以控制雷达以及其他模块的工作。上位机具有多个雷达接口,雷达可通过传输线缆连接至雷达接口,从而将雷达接入上位机中,以便使上位机可以接收雷达的数据,并对雷达进行控制。
本发明实施例的点云数据的存储方法、回放方法及计算机可读介质,以包括文件层、数据帧层和数据包层的至少三层结构的方式存储点云数据,其允许用户在播放点云时即可进行在线储存,存储的点云文件具有较强的鲁棒性,即使应用层程序在运行中崩溃,残存的点云文件仍然能够播放;并且,根据该存储方法存储的点云文件只需很短的解析时间即可进行回放。
图1示出了根据本发明的一个实施例的、用于点云数据的存储方法100的流程图,存储方法100可以用于将从网络层获取的点云数据以文件的形式进行转存,以供后续的算法相关人员进行研发,或者通过特定软件进行数据回放。所述存储方法采用至少三层结构的方式进行存储,所述三层结构包括文件层、数据帧层和数据包层。
本发明实施例中,将根据存储方法100存储所得的点云文件的格式命名为.lvx文件,Lvx文件的层级结构如图2所示,该Lvx文件包括文件层201、数据帧层202和数据包层203。
如图1所示,用于点云数据的存储方法100包括如下步骤:
在步骤S110,获取数据帧,存储于所述文件层201,所述文件层201包括文件头、设备信息表和至少一个所述数据帧。
参见图2,文件层201中存储有文件头、设备信息表和至少一个所述数据帧。在存储文件层201时,除了必要的文件头和设备信息表 之外,将点云数据以数据帧的形式进行组织,从而便于回放。示例性地,可以以20Hz的频率储存所有的数据帧,也即单帧的实际持续时间为50ms。
在一个实施例中,至少获取文件标志、协议版本信息和校验位,并作为所述文件头存储于所述文件层201。作为示例,文件头中包含的数据名称、偏移量及数据类型如表1所示:
表1
其中,所述magic code(幻码)即所述文件标识,用于区分文件类型,也用来标识文件的起始码,其数据类型可以是有符号整型(int8_t),占16个字节。所述version(版本)代表所述协议版本信息,其数据类型可以是无符号整型(uint8_t),偏移量为16个字节,占4个字节,其可以表示为,例如,version[0]=AA,version[1]=BB,version[2]=CC,version[3]=DD,等等。crc32代表所述文件头的校验和,用于实现对数据正确性的检验,当然,除此之外,也可以采用其他类型的校验手段。
在一个实施例中,获取所述点云数据所隶属的传感器信息,并存储于设备信息表中。进一步地,获取所述点云数据所隶属的传感器的数目和每个传感器的设备信息,以存储于所述设备信息表中。根据设备信息表中的信息,用户可以获取点云文件录制时所使用的传感器的设备信息。
示例性地,所述设备信息表中包含的数据名称、偏移量及数据类型如表2所示:
表2
其中,info count(information count,信息数目)代表该点云文件中的点云数据所隶属的传感器的数目。生成点云数据的传感器用于感测外部环境信息,例如,环境目标的距离信息、方位信息、反射强度信息、速度信息等,其可以为激光雷达,也可以为其它雷达或者测距装置。当应用于多雷达系统时,所述传感器的数目可以为多个,多个传感器可分布在可移动平台的不同的位置,用于检测可移动平台周围不同位置/方向的物体信息。
Device info list(设备信息列表)中存储有每个传感器的设备信息。当传感器数目为N时,设备信息列表中包括N个传感器的信息,其中每个传感器的设备信息如表3所示:
表3
参照表3,至少获取每个传感器的序列号、中心板序列号、设备索引、设备类型、以及传感器的角度和坐标,并作为设备信息存储于所述设备列表。其中:
当所述传感器为激光雷达时,所述传感器序列号包括雷达序列号(lidar sn code),具体包括雷达广播码(LiDAR Broadcast Code)。该传感器序列号可用于表示点云数据所隶属的传感器,也即生成点云数据的传感器。
中心板序列号(hub sn code)为与每个传感器连接的中心板的序列号。所述中心板包括中控电路数据汇总,其起到了数据汇总和转发的作用,可以将来自上位机的信息转发给每个传感器,也可以将来自传感器的信息发送到上位机,中心板具有时间对齐的功能,可以将多个传感器的数据在时间上对齐。然而,需要注意的是,传感器可以直接与上位机进行信息交互,也可以通过中心板与所述上位机进行信息交互。示例性地,若未检测到所述传感器与所述中心板连接,则将所述中心板的序列号设为空。在其他实施例中,对于已确定不包括中心板的雷达系统,所述设备信息表中也可以不包括有关于中心板的信息。
设备索引(device index)为当前传感器在所述设备信息列表中的索引。其与数据包头中的索引对应,可用于区分每个数据包所隶属的传感器,便于用户针对每个传感器进行点云分类。
Device type(设备类型)中包含有传感器设备类型信息。所述设备类型信息包括传感器类型或中心板类型。示例性地,0代表中心板,1,2,3分别代表三种不同的雷达类型。
示例性地,传感器的角度信息包括传感器的翻滚角(roll)、俯仰角(pitch)和偏航角(yaw);传感器的位置信息包括传感器的x、y、z轴坐标。传感器的角度和位置信息的信息类型可以为浮点型(float)。
需要说明的是,设备信息列表中的上述信息并非都是必需的,并且,其顺序只是示例性的。或者,除了上述信息之外,所述设备信息列表中也可以存储有其他传感器信息。
在步骤S120中,获取每帧数据,存储于所述数据帧层202,所述数据帧层202包含所述文件层201中每个所述数据帧的信息。
其中,获取当前数据帧的一些必要信息,以作为帧头(Frame Header)存储,作为示例,每个帧头中存储的信息如表4所示:
表4
在表4中,current offset(当前偏移地址)即当前帧的起始偏移地址,next offset(下一偏移地址)即下一帧的起始偏移地址。由于帧头中包含有当前帧的偏移地址和下一帧的偏移地址,所有数据帧的结构类似于链表,用户可以从当前帧的帧头中得知当前帧的信息和下一帧的信息,以便于快速遍历整个文件。
Frame index(帧索引)即当前帧的索引。具体地,获取当前帧的索引,以存储于所述帧头中。使用帧索引可用于快速访问特定的数据帧。
Pack cnt(package count,数据包数目),用于表示当前帧中包含的数据包的数目。示例性地,数据帧的存储频率为20Hz,即每一帧在的数据所对应的实际时间为50ms。然而由于实际录制时应用层的定时器精度的波动,因此实际上每一帧中数据包的数量未必相等,可能会存在一定的波动。
在步骤S130,获取数据帧中的数据包,存储于所述数据包层203,所述数据包层203包含所述数据帧层202中每个所述数据包的信息,其中,每个所述数据包包括数据包头和点云数据,所述数据包头中包含有当前数据包所隶属的传感器的设备信息和时间戳信息。所述数据包层中包含有大量的点云数据,用户可以通过数据包内的数据,应用自行开发的算法,完成更多的操作。
作为示例,每一个数据包中包含的数据名称、偏移量及数据类型如表5所示:
表5
具体地,获取设备索引(device index),以存储于所述数据包头中。所述设备索引与设备信息列表中的索引一一映射,用于区分当前数据包所隶属的传感器。即,每个数据帧中的多个数据包可分别属于不同的传感器,但每个数据包中的点云数据来自同一传感器。
示例性地,所述方法还包括获取协议版本信息(version),以存储于所述数据包头中。所述协议版本为数据包协议版本(Package Protocol Version)。
示例性地,所述方法还包括获取连接所述传感器的端口地址以及传感器ID,以存储于所述数据包头中。
其中,所述端口地址即表5中所示的slot id(端口id),其代表用于连接传感器的端口的地址。示例性地,若传感器不经过中心板,直接与上位机连接,则slot id默认为0,若传感器连接中心板,中心板与上位机连接,则slot id为中心板端口号。
当传感器为雷达时,所述传感器ID为雷达ID(LiDAR id)。其中,单体雷达的序号为0,三合一雷达的编号分别为0,1,2。RSVD(Reserved)为调试预留位。Status code(状态码)用于表示传感器在生成点云数据时的状态。
示例性地,所述方法还包括获取时间戳类型(timestamp type)和时间戳(timestamp),以存储于所述数据包头中。在数据包中存储一个时间戳,用于表示数据包中第一个点的时间。用户可根据所述时间戳定位到特定的数据包进行回放。所述时间戳类型代表时间戳的种类。示例性地,所述时间戳的单位为纳秒(Nanosecond),或者,所述时间戳的格式为协调世界时格式(UTC Format)。
示例性地,所述方法还包括获取数据类型(data type)信息,以存储于所述数据包头中。所述数据类型信息用于表明原始点云数据的坐标格式。示例性地,所述数据类型信息包括笛卡尔坐标(Cartesian Coordinate)类型或球坐标(Spherical Coordinate)类型。例如,可以 用0来表示笛卡尔坐标类型,用1来表示球坐标类型。
需要说明的是,数据包头中的上述信息并非都是必需的。或者,除了上述信息之外,所述数据包头中也可以存储有其他关于所述数据包的信息。
在一个实施例中,应用层程序从网络层获取的点云数据中,每个数据包中包含有100个点的点云数据。因此相对应地,在根据本发明实施例的存储方法100中,在每个所述数据包中存储100个点的点云数据,每个点的点云数据如表6所示:
表6
如表6所示,所述每个点的点云数据包括每个点的坐标和亮度值(intensity)。其中,每个点的坐标为经过外参计算的X轴、Y轴和Z轴坐标,其数据类型为浮点型。所述亮度值可以表征待测物体在该点处的反射率,其数据类型可以是无符号整型。基于大量的点云数据可以重建被测对象相对传感器的空间三维信息。
基于上面的描述,根据本发明实施例的点云数据的存储方法允许用户在播放点云时即可进行在线储存,存储的点云文件具有较强的鲁棒性,即使应用层程序在运行中崩溃,残存的点云文件仍然能够播放;并且,根据该存储方法存储的点云文件只需很短的解析时间即可进行回放。
此外,根据本发明实施例,还提供了一种计算机可读介质,在所述计算机可读介质上存储了程序指令,在所述程序指令被计算机或处理器运行时用于执行本发明实施例的用于点云数据的存储方法100的相应步骤。所述计算机可读介质例如可以包括智能电话的存储卡、平板电脑的存储部件、个人计算机的硬盘、只读存储器(ROM)、可擦除可编程只读存储器(EPROM)、便携式紧致盘只读存储器 (CD-ROM)、USB存储器、或者上述计算机可读介质的任意组合。所述计算机可读介质可以是一个或多个计算机可读介质的任意组合。
本发明实施例另一方面提供了一种点云文件的回放方法,用于回放根据上述存储方法所存储的点云文件。如上所述,所述帧头中包含有数据帧的地址信息,具体包括当前帧的起始偏移地址和下一帧的起始偏移地址以及帧索引,用户通过每一帧的帧头信息对点云文件进行遍历,即可以得到总帧数以及每一帧在文件中的偏移地址。在对点云文件进行回放时,用户可以根据所述帧头中包含的所述地址信息定位到特定的数据帧进行回放。在另一实施例中,由于每个数据包中均包含时间戳,在对点云文件进行回放时,用户可以根据所述数据包头中包含的时间戳信息定位到特定的数据包进行回放。所述点云文件的具体结构可参见上文,为了简洁,在此不再赘述。
基于上面的描述,根据本发明实施例的点云文件的回放方法可快速回放点云文件,允许用户对文件进行解析之后还原成原始点云数据并进行算法分析。
此外,根据本发明实施例,还提供了一种计算机可读介质,在所述计算机可读介质上存储了程序指令,在所述程序指令被计算机或处理器运行时用于执行本发明实施例的用于点云文件的回放方法的相应步骤。所述计算机可读介质例如可以包括智能电话的存储卡、平板电脑的存储部件、个人计算机的硬盘、只读存储器(ROM)、可擦除可编程只读存储器(EPROM)、便携式紧致盘只读存储器(CD-ROM)、USB存储器、或者上述计算机可读介质的任意组合。所述计算机可读介质可以是一个或多个计算机可读介质的任意组合。
尽管这里已经参考附图描述了示例实施例,应理解上述示例实施例仅仅是示例性的,并且不意图将本发明的范围限制于此。本领域普通技术人员可以在其中进行各种改变和修改,而不偏离本发明的范围和精神。所有这些改变和修改意在被包括在所附权利要求所要求的本发明的范围之内。
本领域普通技术人员可以意识到,结合本文中所公开的实施例描述的各示例的单元及算法步骤,能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行, 取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能,但是这种实现不应认为超出本发明的范围。
在本申请所提供的几个实施例中,应该理解到,所揭露的设备和方法,可以通过其它的方式实现。例如,以上所描述的设备实施例仅仅是示意性的,例如,所述单元的划分,仅仅为一种逻辑功能划分,实际实现时可以有另外的划分方式,例如多个单元或组件可以结合或者可以集成到另一个设备,或一些特征可以忽略,或不执行。
在此处所提供的说明书中,说明了大量具体细节。然而,能够理解,本发明的实施例可以在没有这些具体细节的情况下实践。在一些实例中,并未详细示出公知的方法、结构和技术,以便不模糊对本说明书的理解。
类似地,应当理解,为了精简本发明并帮助理解各个发明方面中的一个或多个,在对本发明的示例性实施例的描述中,本发明的各个特征有时被一起分组到单个实施例、图、或者对其的描述中。然而,并不应将该本发明的方法解释成反映如下意图:即所要求保护的本发明要求比在每个权利要求中所明确记载的特征更多的特征。更确切地说,如相应的权利要求书所反映的那样,其发明点在于可以用少于某个公开的单个实施例的所有特征的特征来解决相应的技术问题。因此,遵循具体实施方式的权利要求书由此明确地并入该具体实施方式,其中每个权利要求本身都作为本发明的单独实施例。
本领域的技术人员可以理解,除了特征之间相互排斥之外,可以采用任何组合对本说明书(包括伴随的权利要求、摘要和附图)中公开的所有特征以及如此公开的任何方法或者设备的所有过程或单元进行组合。除非另外明确陈述,本说明书(包括伴随的权利要求、摘要和附图)中公开的每个特征可以由提供相同、等同或相似目的替代特征来代替。
此外,本领域的技术人员能够理解,尽管在此所述的一些实施例包括其它实施例中所包括的某些特征而不是其它特征,但是不同实施例的特征的组合意味着处于本发明的范围之内并且形成不同的实施例。例如,在权利要求书中,所要求保护的实施例的任意之一都可以 以任意的组合方式来使用。
本发明的各个部件实施例可以以硬件实现,或者以在一个或者多个处理器上运行的软件模块实现,或者以它们的组合实现。本领域的技术人员应当理解,可以在实践中使用微处理器或者数字信号处理器(DSP)来实现根据本发明实施例的一些模块的一些或者全部功能。本发明还可以实现为用于执行这里所描述的方法的一部分或者全部的装置程序(例如,计算机程序和计算机程序产品)。这样的实现本发明的程序可以存储在计算机可读介质上,或者可以具有一个或者多个信号的形式。这样的信号可以从因特网网站上下载得到,或者在载体信号上提供,或者以任何其他形式提供。
应该注意的是上述实施例对本发明进行说明而不是对本发明进行限制,并且本领域技术人员在不脱离所附权利要求的范围的情况下可设计出替换实施例。在权利要求中,不应将位于括号之间的任何参考符号构造成对权利要求的限制。本发明可以借助于包括有若干不同元件的硬件以及借助于适当编程的计算机来实现。在列举了若干装置的单元权利要求中,这些装置中的若干个可以是通过同一个硬件项来具体体现。单词第一、第二、以及第三等的使用不表示任何顺序。可将这些单词解释为名称。
以上所述,仅为本发明的具体实施方式或对具体实施方式的说明,本发明的保护范围并不局限于此,任何熟悉本技术领域的技术人员在本发明揭露的技术范围内,可轻易想到变化或替换,都应涵盖在本发明的保护范围之内。本发明的保护范围应以权利要求的保护范围为准。
Claims (30)
- 一种用于点云数据的存储方法,其特征在于,采用至少三层结构的方式进行存储,所述三层结构包括文件层、数据帧层和数据包层,其中所述存储方法包括:获取数据帧,存储于所述文件层,所述文件层包括文件头、设备信息表和至少一个所述数据帧;获取每帧数据,存储于所述数据帧层,所述数据帧层包含所述文件层中每个所述数据帧的信息,其中,每个所述数据帧包括帧头和至少一个数据包,所述帧头中包含有数据帧的地址信息;获取数据帧中的数据包,存储于所述数据包层,所述数据包层包含所述数据帧层中每个所述数据包的信息,其中,每个所述数据包包括数据包头和点云数据,所述数据包头中包含有当前数据包所隶属的传感器的设备信息和时间戳信息。
- 如权利要求1所述的存储方法,其特征在于,至少获取文件标志、协议版本信息和校验位,并作为所述文件头存储于所述文件层。
- 如权利要求1所述的存储方法,其特征在于,获取所述点云数据所隶属的传感器信息,并存储于所述设备信息表中,所述设备信息表中包含所述点云数据所隶属的传感器的数目和每个传感器的设备信息。
- 如权利要求3所述的存储方法,其特征在于,将所述每个传感器的设备信息存储于所述设备信息表中,包括:获取该传感器在所述设备信息表中的索引并作为设备信息进行存储。
- 如权利要求3所述的存储方法,其特征在于,将所述每个传感器的设备信息存储于所述设备信息表中,包括:获取该传感器的序列号并作为设备信息进行存储。
- 如权利要求3所述的存储方法,其特征在于,包括:获取与所述每个传感器连接的中心板的序列号,并将所述中心板的序列号存储于所述设备信息中。
- 如权利要求6所述的存储方法,其特征在于,若未检测到所述传感器与所述中心板连接,则将所述中心板的序列号设为空。
- 如权利要求3所述的存储方法,其特征在于,包括:获取所 述每个传感器的设备类型信息,并存储在对应的设备信息中。
- 如权利要求8所述的存储方法,其特征在于,所述设备类型信息包括传感器类型或中心板类型。
- 如权利要求3所述的存储方法,其特征在于,包括:获取所述每个传感器的角度和坐标,并存储在对应的设备信息中。
- 如权利要求1所述的存储方法,其特征在于,所述地址信息包括当前帧的起始偏移地址和下一帧的起始偏移地址。
- 如权利要求1所述的存储方法,其特征在于,还包括:获取当前帧的索引,并存储于所述帧头中。
- 如权利要求1所述的存储方法,其特征在于,还包括:获取当前帧中所述数据包的数目,并存储于所述帧头中。
- 如权利要求1所述的存储方法,其特征在于,所述数据帧的存储频率为20Hz。
- 如权利要求1所述的存储方法,其特征在于,还包括:在所述数据包头中存储与所述设备信息表相对应的设备索引,用于标记当前数据包所隶属的传感器。
- 如权利要求1所述的存储方法,其特征在于,还包括:获取连接所述传感器的端口地址,并存储于所述数据包头中。
- 如权利要求1所述的存储方法,其特征在于,还包括:获取传感器ID,并存储于所述数据包头中,所述传感器ID用于区分单体传感器和三合一传感器。
- 如权利要求1所述的存储方法,其特征在于,还包括:获取数据包协议版本信息,并存储于所述数据包头中。
- 如权利要求1所述的存储方法,其特征在于,还包括:在所述数据包头预留调试预留位。
- 如权利要求1所述的存储方法,其特征在于,还包括:获取传感器状态信息,并存储于所述数据包头中。
- 如权利要求1所述的存储方法,其特征在于,所述时间戳信息包括所述数据包的时间戳类型和时间戳。
- 如权利要求21所述的存储方法,其特征在于,所述时间戳的单位为纳秒,或者,所述时间戳的格式为协调世界时。
- 如权利要求1所述的存储方法,其特征在于,还包括获取数据类型信息,并存储于所述数据包头中,所述数据类型信息用于表明原始点云数据的坐标格式。
- 如权利要求23所述的存储方法,其特征在于,所述数据类型信息包括笛卡尔坐标类型或球坐标类型。
- 如权利要求1所述的存储方法,其特征在于,所述点云数据包括100个点的点云数据。
- 如权利要求1所述的存储方法,其特征在于,所述点云数据包括每个点的坐标和亮度值。
- 如权利要求1-26之一所述的存储方法,其特征在于,所述传感器包括激光雷达。
- 一种点云文件的回放方法,用于回放根据权利要求1-27之一所述存储方法所存储的点云文件,其特征在于,所述回放方法包括:根据所述帧头中包含的所述地址信息定位到特定的数据帧进行回放,或者,根据所述数据包头中包含的时间戳信息定位到特定的数据包进行回放。
- 一种计算机可读介质,所述计算机可读介质上存储有计算机程序,所述计算机程序在运行时执行如权利要求1-27之一所述的用于点云数据的存储方法。
- 一种计算机可读介质,所述计算机可读介质上存储有计算机程序,所述计算机程序在运行时执行如权利要求28所述的点云文件的回放方法。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201980005533.XA CN111771196A (zh) | 2019-01-30 | 2019-01-30 | 用于点云数据的存储方法、回放方法及计算机可读介质 |
PCT/CN2019/073984 WO2020154975A1 (zh) | 2019-01-30 | 2019-01-30 | 用于点云数据的存储方法、回放方法及计算机可读介质 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2019/073984 WO2020154975A1 (zh) | 2019-01-30 | 2019-01-30 | 用于点云数据的存储方法、回放方法及计算机可读介质 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2020154975A1 true WO2020154975A1 (zh) | 2020-08-06 |
Family
ID=71840771
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2019/073984 WO2020154975A1 (zh) | 2019-01-30 | 2019-01-30 | 用于点云数据的存储方法、回放方法及计算机可读介质 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN111771196A (zh) |
WO (1) | WO2020154975A1 (zh) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112948332B (zh) * | 2021-03-04 | 2023-07-21 | 北京奇艺世纪科技有限公司 | 一种点云文件封装、解析方法及装置 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101170345A (zh) * | 2007-11-30 | 2008-04-30 | 航天东方红卫星有限公司 | 一种卫星分包遥测数据的解包方法 |
CN104750854A (zh) * | 2015-04-16 | 2015-07-01 | 武汉海达数云技术有限公司 | 一种海量三维激光点云压缩存储及快速加载显示方法 |
CN106991143A (zh) * | 2017-03-22 | 2017-07-28 | 苏州中科图新网络科技有限公司 | 多层级图像文件、生成方法及装置、读取方法及装置 |
CN108234552A (zh) * | 2016-12-15 | 2018-06-29 | 腾讯科技(深圳)有限公司 | 一种数据存储方法及装置 |
-
2019
- 2019-01-30 CN CN201980005533.XA patent/CN111771196A/zh active Pending
- 2019-01-30 WO PCT/CN2019/073984 patent/WO2020154975A1/zh active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101170345A (zh) * | 2007-11-30 | 2008-04-30 | 航天东方红卫星有限公司 | 一种卫星分包遥测数据的解包方法 |
CN104750854A (zh) * | 2015-04-16 | 2015-07-01 | 武汉海达数云技术有限公司 | 一种海量三维激光点云压缩存储及快速加载显示方法 |
CN108234552A (zh) * | 2016-12-15 | 2018-06-29 | 腾讯科技(深圳)有限公司 | 一种数据存储方法及装置 |
CN106991143A (zh) * | 2017-03-22 | 2017-07-28 | 苏州中科图新网络科技有限公司 | 多层级图像文件、生成方法及装置、读取方法及装置 |
Non-Patent Citations (2)
Title |
---|
SHAO, ZHUXING ET AL: "Large File Storage Solution for OpenStack-based Cloud Storage System", COMPUTER ENGINEERING AND DESIGN, vol. 36, no. 2, 28 February 2015 (2015-02-28), ISSN: 1000-7024 * |
YANG, JIANSI ET AL: "A Management Method Base on Big File Data Storage Structure and Multi Versions for Remote Sensing Image", GEOMATICS WORLD, vol. 21, no. 2, 30 April 2014 (2014-04-30), ISSN: 1672-1586 * |
Also Published As
Publication number | Publication date |
---|---|
CN111771196A (zh) | 2020-10-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20220036648A1 (en) | Method and terminal device for determining occluded area of virtual object | |
CN110401715B (zh) | 资源收集任务管理方法、装置、存储介质及系统 | |
CN106774844B (zh) | 一种用于虚拟定位的方法及设备 | |
CN110648363B (zh) | 相机姿态确定方法、装置、存储介质及电子设备 | |
US20210182633A1 (en) | Localization method and helmet and computer readable storage medium using the same | |
CN111311757B (zh) | 一种场景合成方法、装置、存储介质及移动终端 | |
JP2019072478A (ja) | 情報処理方法及び装置、記憶媒体、電子機器 | |
US20140295891A1 (en) | Method, server and terminal for information interaction | |
CN111339086A (zh) | 区块处理方法、基于区块链的数据查询方法及装置 | |
CN109255749A (zh) | 自主和非自主平台中的地图构建优化 | |
US11651569B2 (en) | System and method for mapping | |
US20200242850A1 (en) | Method and apparatus for displaying with 3d parallax effect | |
JP7535177B2 (ja) | 加速レイトレーシングのシステムおよび方法 | |
CN110061892A (zh) | 实时显示网络状况的方法、系统、存储介质、及电子设备 | |
CN113840637A (zh) | 共享坐标空间内的共享增强现实游戏 | |
US20220109617A1 (en) | Latency determinations for human interface devices | |
US12079920B2 (en) | System and method for accelerated ray tracing with asynchronous operation and ray transformation | |
CN113343457A (zh) | 自动驾驶的仿真测试方法、装置、设备及存储介质 | |
CN106643518A (zh) | 一种利用双目摄像装置测量距离和大小的方法和装置 | |
WO2020154975A1 (zh) | 用于点云数据的存储方法、回放方法及计算机可读介质 | |
CN109559382A (zh) | 智能导游方法、装置、终端和介质 | |
US20200042821A1 (en) | Interactive method and interactive system | |
CN110191316A (zh) | 一种信息处理方法及装置、设备、存储介质 | |
WO2022183906A1 (zh) | 成像方法、装置、设备及存储介质 | |
CN108985831A (zh) | 一种线下交易的判别方法、装置、及计算机设备 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 19912925 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 19912925 Country of ref document: EP Kind code of ref document: A1 |