WO2021226780A1 - 一种实景地图生成方法、装置、设备及可读存储介质 - Google Patents

一种实景地图生成方法、装置、设备及可读存储介质 Download PDF

Info

Publication number
WO2021226780A1
WO2021226780A1 PCT/CN2020/089553 CN2020089553W WO2021226780A1 WO 2021226780 A1 WO2021226780 A1 WO 2021226780A1 CN 2020089553 W CN2020089553 W CN 2020089553W WO 2021226780 A1 WO2021226780 A1 WO 2021226780A1
Authority
WO
WIPO (PCT)
Prior art keywords
real
scene
image
map
preset
Prior art date
Application number
PCT/CN2020/089553
Other languages
English (en)
French (fr)
Inventor
陈尊裕
吴沛谦
张仲文
吴珏其
胡斯洋
陈欣
Original Assignee
蜂图志科技控股有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 蜂图志科技控股有限公司 filed Critical 蜂图志科技控股有限公司
Priority to CN202080001086.3A priority Critical patent/CN111801664A/zh
Priority to PCT/CN2020/089553 priority patent/WO2021226780A1/zh
Publication of WO2021226780A1 publication Critical patent/WO2021226780A1/zh

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
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/29—Geographical information databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
    • G06F16/73—Querying
    • G06F16/732—Query formulation
    • G06F16/7328—Query by example, e.g. a complete video frame or video sequence
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
    • G06F16/78—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/783—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G06F16/7847—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using low-level visual features of the video content
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
    • G06F16/78—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/787—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using geographical or spatial information, e.g. location

Definitions

  • the present invention relates to the field of real map technology, in particular to a real map generating method, a real map generating device, a real map generating device and a computer readable storage medium.
  • the real image is an image that records the real environment, and the real image with geographic location information is often called a street view map in an outdoor environment, and a real map in an indoor environment.
  • GPS Global Positioning System, Global Positioning System
  • other satellite positioning signals are weak, and it is impossible to obtain accurate coordinate information of each image in the entire real-world map by using satellite positioning as in an outdoor open environment. Therefore, when generating a real-world map, related technologies mostly use indoor positioning technology to obtain coordinate information of the real-world map.
  • indoor positioning technology requires the deployment of indoor positioning equipment in advance, and requires high on-site scenes in the process of generating real-world maps. In most cases, the site needs to be cleared to prevent interference with signals and inaccurate real-world maps. Therefore, the related technology consumes a lot of manpower and time when generating the real-world map, and the real-world map generation efficiency is low.
  • the purpose of the present invention is to provide a real-scene map generation method, real-scene map generation device, real-scene map generation equipment and computer readable storage medium, which solves the need for related technologies to consume a lot of manpower and time, and the real-scene map generation efficiency is low.
  • the problem is to provide a real-scene map generation method, real-scene map generation device, real-scene map generation equipment and computer readable storage medium, which solves the need for related technologies to consume a lot of manpower and time, and the real-scene map generation efficiency is low.
  • the present invention provides a method for generating a real scene map, including:
  • a real-world map is generated using the real-world image set and the target coordinate set.
  • the acquiring the real-scene image set and the anchor point coordinates corresponding to the real-scene image set includes:
  • the acquiring the real-scene image set and the anchor point coordinates corresponding to the real-scene image set includes:
  • the calculating a preset coordinate set corresponding to the real-scene image set by using the anchor point coordinates includes:
  • the preset coordinate set is constructed using each of the preset coordinates.
  • the performing motion restoration structure processing on the preset coordinate set according to the real-scene image set to obtain the target coordinate set includes:
  • the performing coordinate correction on the preset coordinate set according to the spatial structure to obtain the target coordinate set includes:
  • the generating a real-world map using the real-world image set and the target coordinate set includes:
  • the present invention also provides a real-scene map generating device, including:
  • An acquiring module configured to acquire a set of real scene images and anchor point coordinates corresponding to the set of real scene images
  • a calculation module configured to use the anchor point coordinates to calculate a preset coordinate set corresponding to the real scene image set
  • a processing module configured to perform motion restoration structure processing on the preset coordinate set according to the real image set to obtain a target coordinate set
  • the generating module is used to generate a real-world map by using the real-scene image set and the target coordinate set.
  • the present invention also provides a real-scene map generating device, including a memory and a processor, wherein:
  • the memory is used to store computer programs
  • the processor is configured to execute the computer program to realize the above-mentioned method for generating a real scene map.
  • the present invention also provides a computer-readable storage medium for storing a computer program, wherein the computer program is executed by a processor to realize the above-mentioned real-scene map generation method.
  • the real scene map generation method acquires a real scene image set and the anchor point coordinates corresponding to the real scene image set; uses the anchor point coordinates to calculate the preset coordinate set corresponding to the real scene image set; and moves the preset coordinate set according to the real scene image set Restoring the structure processing to obtain the target coordinate set; use the real scene image set and the target coordinate set to generate the real scene map.
  • this method acquires the corresponding anchor point coordinates at the same time when acquiring the real-scene image set, and the preset coordinate set corresponding to the real-scene image set can be calculated through the anchor point coordinates.
  • the preset coordinate set includes the preset coordinates of each image in the real-scene image set. , That is, approximate coordinates.
  • the motion restoration structure processing is performed on the preset coordinate set, and the preset coordinate set can be adjusted according to the real image in the real image set to obtain the accurate coordinate set corresponding to the real image set, that is, the target coordinate set.
  • the motion recovery structure processing based on the real image set can ensure the accuracy of the target coordinate set.
  • Using the target coordinate set and the real image set to generate a real map can save the installation of indoor positioning equipment and use indoors on the basis of ensuring the accuracy of the real map.
  • the large amount of manpower and time required by the positioning device to generate the map improves the generation efficiency of the real map, and solves the problem that the related technology requires a lot of manpower and time, and the generation of the real map is low in efficiency.
  • the present invention also provides a real-scene map generation device, real-scene map generation equipment, and computer-readable storage medium, which also have the above-mentioned beneficial effects.
  • FIG. 1 is a flowchart of a method for generating a real-world map according to an embodiment of the present invention
  • FIG. 2 is a flowchart of a method for acquiring a set of real scene images and anchor point coordinates according to an embodiment of the present invention
  • FIG. 3 is a flow chart of another method for acquiring real scene image sets and anchor point coordinates according to an embodiment of the present invention
  • FIG. 4 is a flowchart of a specific method for generating a real-world map provided by an embodiment of the present invention
  • FIG. 5 is a schematic structural diagram of a device for generating a real-world map according to an embodiment of the present invention.
  • Fig. 6 is a schematic structural diagram of a real-scene map generating device provided by an embodiment of the present invention.
  • the real image is an image that records the real environment, and the real image with geographic location information is often called a street view map in an outdoor environment, and a real map in an indoor environment.
  • Most of the real-world map construction schemes focus on the outdoor environment, that is, they are used to generate street-view images.
  • real-life photos are collected by on-board equipment or drones.
  • the geographic location information corresponding to the real-world photos is generally through vehicles or vehicles equipped with on-board equipment. Obtained by UAV GPS and other satellite positioning equipment.
  • a positioning terminal In an indoor environment, the signal of satellite positioning equipment such as GPS will be weakened, which will lead to the problem of inaccurate positioning.
  • related technologies generally use indoor positioning technology to generate the real map.
  • a positioning terminal is placed indoors where a real scene map needs to be drawn, and the coordinates corresponding to the positioning terminal are accurate geographic coordinates.
  • the positioning terminal uses its own geographic coordinates and the position relationship with the image acquisition device to locate the image acquisition device and obtain the corresponding image. Accurate geographic location information to complete the generation of real-world maps.
  • the present application provides a method for generating a real scene map, as well as a corresponding device, equipment, and computer-readable storage medium.
  • the method does not need to set up an indoor positioning terminal, and by performing motion restoration structure processing on the coordinate set corresponding to the real image set, the coordinates can be corrected based on the images in the real image set to ensure the accuracy of the target coordinate set corresponding to the real image set.
  • the consumption of manpower and time is reduced, and the efficiency is greatly improved.
  • FIG. 1 is a flowchart of a method for generating a real-world map according to an embodiment of the present invention, and the method includes:
  • S101 Acquire a set of real scene images and anchor point coordinates corresponding to the set of real scene images.
  • the real-scene image set includes multiple real-scene images, the real-scene images are used to generate real-scene maps, and the specific number of real-scene images is not limited in this embodiment.
  • the real scene image is a continuous image, that is, there are overlapping parts between adjacent images, so that the motion recovery structure processing can be performed on it later.
  • the method for acquiring a set of real scene images is not limited in this embodiment.
  • a preset path may be set, and real scene images are acquired according to the preset path, and then a real scene image set is formed.
  • the preset path may be appropriately deviated; or the fixed preset path may not be set, and the actual path may be selected according to actual needs to obtain the real scene image, and then form the real scene image set.
  • the anchor point coordinates correspond to the real scene image set, and the specific number is multiple.
  • the anchor point coordinates are the coordinates corresponding to the target real scene image in the real scene image set. It should be noted that the anchor point coordinates are accurate coordinates obtained by non-indoor positioning technology.
  • the specific acquisition method is not limited in this embodiment. For example, satellite positioning technology may be used to acquire the anchor point coordinates; or a mobile phone or a wireless communication signal base station may be used to locate the anchor point to obtain the anchor point coordinates.
  • the anchor point corresponding to the anchor point coordinate can be any position point, for example, it can be the starting point, the end point, the turning point or other arbitrary points of the preset path.
  • the anchor point coordinates are the coordinates of the target real scene image
  • the anchor point corresponding to the anchor point coordinates must be on the actual path of the real scene image set.
  • this embodiment does not limit the order of determining the preset path and anchor points. Specifically, a preset path may be determined first, and a point with accurate coordinates in the preset path may be selected as an anchor point; or a plurality of anchor points may be determined first, and the preset path may be set based on the anchor points.
  • S102 Calculate a preset coordinate set corresponding to the real scene image set by using the anchor point coordinates.
  • the anchor point coordinates are used to calculate the preset coordinate set corresponding to the real image set.
  • the preset coordinate set includes a plurality of preset coordinates, and each preset coordinate corresponds to each real scene image in the real scene image set.
  • the preset coordinates are the approximate coordinates of the real scene image, which correspond to the real scene image.
  • the specific calculation method is not limited in this embodiment. For example, when there is a preset path, the length between each anchor point in the preset path can be divided equally , Get multiple preset points, and use the anchor point coordinates to calculate the coordinates of the preset points to obtain the preset coordinates, and then form a preset coordinate set; when there is no preset path, you can use a straight line or a curve to connect each anchor The fixed points are connected to obtain a simulated path, the preset points are determined on the simulated path and the preset coordinates are calculated, and the preset coordinate set is finally obtained.
  • S103 Perform motion restoration structure processing on the preset coordinate set according to the real scene image set to obtain the target coordinate set.
  • the coordinates in the preset coordinate set are approximate coordinates, which are inaccurate. In order to generate an accurate real-world map, it needs to be adjusted to obtain the target coordinate set, which is the accurate coordinate set corresponding to the real-world image set.
  • the motion recovery structure is Structure From Motion, SFM, by locating multiple common feature points in different pictures, and according to the principle of straight propagation of light, combining the changes in the pixel positions of the feature points in at least two pictures and image acquisition
  • the parameters of the device are used to calculate the location of the image acquisition device, that is, the target location corresponding to the real image.
  • the parameters of the device may include, but are not limited to, focal length, photosensitive element size, camera matrix, and so on.
  • the motion recovery structure processing can use the SFM algorithm, or other similar algorithms or programs can be used to process the preset coordinate set, and only the target coordinate set needs to be obtained.
  • This embodiment does not limit the representation form of the target coordinates in the target coordinate set, which is the same as the representation form of the preset coordinate set, that is, the representation form of the anchor point coordinates, for example, may be in the form of latitude and longitude coordinates; or other options may be selected according to actual needs.
  • the form for example, can be a coordinate representation form in a custom coordinate system.
  • S104 Generate a real scene map by using the real scene image set and the target coordinate set.
  • the real scene map can be generated using the real scene image set and the target coordinate set.
  • This embodiment does not limit the specific generation method of the real scene map, and can refer to related technologies, and will not be repeated here.
  • the motion recovery structure is used to obtain the target coordinate set, which not only ensures the accuracy of the real map, but also saves the installation of indoor positioning equipment and the use of indoor positioning equipment to generate A lot of manpower and time required for the map.
  • the corresponding anchor point coordinates are obtained at the same time when the real-scene image set is acquired.
  • the anchor point coordinates can be calculated to obtain the preset coordinate set corresponding to the real-scene image set, and the preset coordinate set includes the real scene.
  • the preset coordinates of each image in the image set that is, approximate coordinates.
  • the motion recovery structure processing is performed on the preset coordinate set, and the preset coordinate set can be adjusted according to the images in the real image set to obtain an accurate coordinate set corresponding to the real image set, that is, the target coordinate set.
  • the motion recovery structure processing based on the real image set can ensure the accuracy of the target coordinate set.
  • Using the target coordinate set and the real image set to generate a real map can save the installation of indoor positioning equipment and use indoors on the basis of ensuring the accuracy of the real map.
  • the large amount of manpower and time required by the positioning device to generate the map improves the generation efficiency of the real map, and solves the problem that the related technology requires a lot of manpower and time, and the generation of the real map is low in efficiency.
  • FIG. 2 is a flowchart of a method for acquiring coordinates of a real image set and anchor points according to an embodiment of the present invention, including:
  • S201 Acquire multiple real-scene images, and use the real-scene images to form a real-scene image set.
  • the method of acquiring real-scene images by taking pictures constitutes a real-scene image set.
  • the image acquisition device can be set on the robot, and the robot can use the image acquisition device to acquire the real-life image. Images; or you can use manual methods to obtain real-life images.
  • an image acquisition device with anti-shake function can be used to acquire real scene images; or an image acquisition device with a pan-tilt can be used to acquire real scene images.
  • This embodiment does not deal with the specific content of the image acquisition device. Limited, for example, it can be a panoramic camera, a normal camera, or a mobile phone.
  • This embodiment does not limit the specific method of acquiring real-scene images.
  • real-scene images can be acquired at preset time intervals during exercise, and the motion can be uniform or non-uniform motion; or real-scene images can be acquired at preset distance intervals. After acquiring the real image, use it to compose the real image set.
  • the target real scene image in this embodiment is a real scene image with accurate coordinates, that is, a real scene image shot at an anchor point. Therefore, after acquiring the real image set, determine the target real image in it to obtain the anchor point coordinates.
  • S203 Obtain the image coordinates corresponding to each target real scene image, and determine the image coordinates as anchor point coordinates.
  • the corresponding image coordinates are acquired, and then the image coordinates are the anchor point coordinates.
  • the anchor point can be determined first, and the real scene image taken at the anchor point can be determined as the target real scene image; or the target real scene image can be determined first , That is, to determine all or part of the real scene image with accurate coordinates as the target real scene image, and determine its shooting position as the anchor point, and determine its coordinates as the anchor point coordinates.
  • the real-scene image set may be obtained by extracting image frames.
  • FIG. 3 is a flowchart of another method for acquiring real-world image sets and anchor point coordinates according to an embodiment of the present invention, including:
  • the real-scene image set is acquired by extracting image frames from the real-scene video.
  • the method for acquiring the real-scene video is not limited in this embodiment, and the above-mentioned real-scene image acquisition method can be referred to.
  • S302 Extract multiple image frames from the real scene video according to a preset sampling frequency, and use the image frames to form a real scene image set.
  • multiple image frames may be extracted from the real scene video according to a preset sampling frequency to form a real scene image set.
  • the specific size of the preset sampling frequency can be set according to actual conditions, which is not limited in this embodiment. Since there are multiple image frames in the real scene video, the flexibility of selecting the real scene image set can be improved. It should be noted that the image frame is only a special form of the real image, and the image frame itself is still the real image.
  • sampling rules can be set according to actual needs, and image frames can be extracted from the live video according to the sampling rules, or image frames can be manually selected. After the image frames are extracted, the image frames in the real image set can be added, deleted, or replaced, which is not limited in this embodiment.
  • the target image frame can be determined from it, and the target image frame is the image frame obtained at the anchor point. Since the selection of the image frame is more flexible, the selection of the target image frame can also be more flexible.
  • S304 Obtain image coordinates corresponding to each target image frame, and determine the image coordinates as anchor point coordinates.
  • the target image frame is the target real-life image
  • the coordinates of the target image frame are the anchor point coordinates
  • FIG. 4 is a flowchart of a specific method for generating a real scene map provided by an embodiment of the present invention, including:
  • S401 Determine a coordinate calculation rule corresponding to a set of real scene images.
  • the coordinate calculation rule is used to calculate the preset coordinates.
  • the coordinate calculation rule corresponds to the real image set.
  • This embodiment does not limit the specific content of the coordinate calculation rule. Specifically, it can be based on the acquisition method of the real image set Determine the coordinate calculation rules. For example, when there is a preset path, the coordinate calculation rules can be set according to the preset path. In some possible implementations, the preset path is appropriately deviated when the real scene image is obtained. In this case, the preset path can still be set Coordinate calculation rules, or coordinate calculation rules can be set according to a preset path and offset.
  • a straight line or a curve can be used to connect the anchor points to obtain a simulated path, so that the coordinate calculation rules can be set according to the simulated path.
  • the preset acquisition interval can also be used to determine the distribution of each real scene image, so as to calculate each preset coordinate.
  • the preset acquisition interval may be a time interval, such as one second; or may be a distance interval, such as 50 cm.
  • S402 Calculate the preset coordinates corresponding to each real scene image in the real scene image set by using the anchor point coordinates and according to the coordinate calculation rule.
  • the preset coordinates corresponding to each real image in the real image set can be obtained.
  • the process of motion recovery structure processing includes three steps S404, S405, and S406, specifically:
  • S404 Perform feature point extraction on the real scene images in the real scene image set, and match the feature points to obtain multiple feature point pairs.
  • the spatial visual structure is a sparse point cloud. It should be noted that, because the feature points in the feature point pair are the same feature in different real-scene images, the multiple real-scene images are adjacent real-scene images.
  • S405 Use the feature point pair to perform a spatial structure restoration operation to obtain a spatial structure.
  • the spatial structure is the spatial structure recorded by the real image, and can also be called a sparse point cloud.
  • the spatial structure restoration operation can be used to determine the relative spatial position between each feature point, and the relative spatial position is the spatial structure corresponding to the feature point.
  • S406 Perform coordinate correction on the preset coordinate set according to the spatial structure to obtain the target coordinate set.
  • coordinate correction is to use the spatial structure to perform coordinate calculation based on a preset coordinate set to obtain the shooting position coordinates corresponding to each real scene image, or called image coordinates, and the shooting position coordinates are the target coordinates in the target coordinate set.
  • step S406 may include:
  • S4061 Perform coordinate correction on the preset coordinate set according to the spatial structure to obtain an intermediate coordinate set.
  • one or more intermediate coordinate sets can be obtained.
  • Different intermediate coordinate sets can be coordinate sets obtained by performing different coordinate corrections on the preset coordinate sets.
  • the number and specific content of this implementation The examples are not limited.
  • S4062 Perform minimizing error processing on the intermediate coordinate set to obtain a target coordinate set.
  • Error minimization processing may also be referred to as a global optimal strategy.
  • the specific processing process is not limited in this embodiment, and related technologies can be referred to.
  • S407 Perform stitching processing on the real scene images in the real scene image set to obtain an initial real scene map.
  • the real scene map is generally the whole image, so after obtaining the target coordinate set, in order to generate the real scene map, the real scene images in the real scene image set need to be combined to obtain the whole real scene map, that is, the initial real scene map.
  • S408 Mark the initial real scene map by using the target coordinate set to obtain the real scene map.
  • the target coordinate set By using the target coordinate set to mark the initial real scene map, it can be accompanied by accurate coordinate information, and finally the real scene map can be obtained.
  • the following describes the real-scene map generation device provided by the embodiment of the present invention.
  • the real-scene map generation device described below and the real-scene map generation method described above can be referred to each other.
  • FIG. 5 is a schematic structural diagram of a real-world map generating apparatus provided by an embodiment of the present invention, including:
  • the obtaining module 510 is used to obtain a real-scene image set and anchor point coordinates corresponding to the real-scene image set;
  • the calculation module 520 is configured to use the anchor point coordinates to calculate a preset coordinate set corresponding to the real scene image set;
  • the processing module 530 is configured to perform motion restoration structure processing on the preset coordinate set according to the real scene image set to obtain the target coordinate set;
  • the generating module 540 is used to generate a real-world map by using the real-world image set and the target coordinate set.
  • the obtaining module 510 includes:
  • the first image set acquiring unit is used to acquire multiple real-scene images, and use the real-scene images to form a real-scene image set;
  • the first determining unit is used to determine the target real-scene image in the real-scene image set
  • the first coordinate acquiring unit is used to acquire the image coordinates corresponding to each target real scene image, and determine the image coordinates as anchor point coordinates.
  • the obtaining module 510 includes:
  • the video acquisition unit is used to acquire the real scene video
  • the second image set acquisition unit is configured to extract multiple image frames from the real scene video according to a preset sampling frequency, and use the image frames to form a real scene image set;
  • the second determining unit is used to determine the target image frame in the real image set
  • the second coordinate acquiring unit is used to acquire the image coordinates corresponding to each target image frame, and determine the image coordinates as anchor point coordinates.
  • the calculation module 520 includes:
  • the calculation rule determination unit is used to determine the coordinate calculation rules corresponding to the real-scene image set
  • the preset coordinate calculation unit is used to use the anchor point coordinates to calculate the preset coordinates corresponding to each image in the real image set according to the coordinate calculation rules;
  • the preset coordinate set construction unit is used to construct a preset coordinate set by using each preset coordinate.
  • processing module 530 includes:
  • the feature point extraction unit is used to extract feature points from the images in the real image set, and to match the feature points to obtain multiple feature point pairs;
  • the spatial structure generating unit is used to perform spatial structure restoration operations using feature point pairs to obtain the spatial structure
  • the coordinate correction unit is used to perform coordinate correction on the preset coordinate set according to the spatial structure to obtain the target coordinate set.
  • the coordinate correction unit includes:
  • the correction subunit is used to perform coordinate correction on the preset coordinate set according to the spatial structure to obtain an intermediate coordinate set
  • the error processing subunit is used for minimizing error processing on the intermediate coordinate set to obtain the target coordinate set.
  • the generating module 540 includes:
  • the stitching unit is used for stitching the images in the real scene image collection to obtain the initial real scene map
  • the marking unit is used to mark the initial real scene map by using the target coordinate set to obtain the real scene map.
  • the corresponding anchor point coordinates are obtained at the same time when the real-scene image set is acquired.
  • the anchor point coordinates can be calculated to obtain the preset coordinate set corresponding to the real-scene image set, and the preset coordinate set includes the real scene
  • the preset coordinates of each image in the image set that is, approximate coordinates.
  • the motion recovery structure processing is performed on the preset coordinate set, and the preset coordinate set can be adjusted according to the images in the real image set to obtain the accurate coordinate set corresponding to the real image set, that is, the target coordinate set.
  • the motion recovery structure processing based on the real image set can ensure the accuracy of the target coordinate set.
  • Using the target coordinate set and the real image set to generate a real map can save the installation of indoor positioning equipment and use indoors on the basis of ensuring the accuracy of the real map.
  • the large amount of manpower and time required by the positioning device to generate the map improves the generation efficiency of the real map, and solves the problem that the related technology requires a lot of manpower and time, and the generation of the real map is low in efficiency.
  • the following describes the real-scene map generation device provided by the embodiment of the present invention.
  • the real-scene map generation device described below and the real-scene map generation method described above can be referred to each other.
  • the real-world map generating device 600 may include a processor 601 and a memory 602, and may further include one or more of a multimedia component 603, an information input/information output (I/O) interface 604, and a communication component 605.
  • a multimedia component 603 may be included in the real-world map generating device 600.
  • I/O information input/information output
  • the processor 601 is used to control the overall operation of the real-world map generating device 600 to complete all or part of the steps in the above-mentioned real-world map generating method; the memory 602 is used to store various types of data to support the real-world map generation device 600
  • the data may include instructions for any application program or method operated on the real-scene map generating device 600, and application-related data.
  • the memory 602 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, SRAM), electrically erasable programmable read-only memory (Electrically erasable programmable read-only memory).
  • EEPROM Erasable Programmable Read-Only Memory
  • EPROM Erasable Programmable Read-Only Memory
  • PROM Programmable Read-Only Memory
  • Read-Only Memory One or more of Only Memory, ROM), magnetic memory, flash memory, magnetic disk or optical disk.
  • the multimedia component 603 may include a screen and an audio component.
  • the screen may be a touch screen, for example, and the audio component is used to output and/or input audio signals.
  • the audio component may include a microphone, which is used to receive external audio signals.
  • the received audio signal may be further stored in the memory 602 or sent through the communication component 605.
  • the audio component also includes at least one speaker for outputting audio signals.
  • the I/O interface 604 provides an interface between the processor 601 and other interface modules.
  • the above-mentioned other interface modules may be keyboards, mice, buttons, and the like. These buttons can be virtual buttons or physical buttons.
  • the communication component 605 is used for wired or wireless communication between the real-view map generating device 600 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination of one or more of them, so the corresponding communication component 605 may include: Wi-Fi components, Bluetooth components, NFC components.
  • the real scene map generation device 600 can be used by one or more application specific integrated circuits (Application Specific Integrated Circuit, ASIC for short), Digital Signal Processor (Digital Signal Processor, DSP), Digital Signal Processing Device (Digital Signal Processing Device, DSPD for short) ), programmable logic device (Programmable Logic Device, PLD), Field Programmable Gate Array (Field Programmable Gate Array, FPGA), controller, microcontroller, microprocessor or other electronic components to implement the above The real-scene map generation method given in the embodiment.
  • ASIC Application Specific Integrated Circuit
  • DSP Digital Signal Processor
  • DSP Digital Signal Processing Device
  • DSPD Digital Signal Processing Device
  • PLD Programmable Logic Device
  • FPGA Field Programmable Gate Array
  • the computer-readable storage medium provided by the embodiment of the present invention will be introduced below.
  • the computer-readable storage medium described below and the method for generating a real-world map described above may correspond to each other and refer to each other.
  • the present invention also provides a computer-readable storage medium with a computer program stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the above-mentioned real-scene map generation method are realized.
  • the computer-readable storage medium may include: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk, etc., which can store program codes Medium.
  • the steps of the method or algorithm described in the embodiments disclosed in this document can be directly implemented by hardware, a software module executed by a processor, or a combination of the two.
  • the software module can be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs, or all areas in the technical field. Any other known storage media.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Multimedia (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Library & Information Science (AREA)
  • Remote Sensing (AREA)
  • Mathematical Physics (AREA)
  • Computational Linguistics (AREA)
  • Processing Or Creating Images (AREA)

Abstract

一种实景地图生成方法、实景地图生成装置、实景地图生成设备及计算机可读存储介质，所述方法包括：获取实景图像集以及与实景图像集对应的锚定点坐标（S101）；利用锚定点坐标计算实景图像集对应的预设坐标集（S102）；根据实景图像集对预设坐标集进行运动恢复结构处理，得到目标坐标集（S103）；利用实景图像集与目标坐标集生成实景地图（S104）；利用基于实景图像集的运动恢复结构处理可以保证目标坐标集的准确性，进而利用目标坐标集与实景图像集生成实景地图，可以在保证实景地图准确性的基础上，节省了设置室内定位设备以及利用室内定位设备生成地图所需的大量人力和时间，提高了实景地图的生成效率。

Description

一种实景地图生成方法、装置、设备及可读存储介质 技术领域
本发明涉及实景地图技术领域，特别涉及一种实景地图生成方法、实景地图生成装置、实景地图生成设备及计算机可读存储介质。
背景技术
实景图像为记录现实环境的图像，而具有地理位置信息的实景图像，在室外环境中，常被称为街景地图，在室内环境中，常被称为实景地图。
在室内环境下，GPS(Global Positioning System，全球定位系统)等卫星定位信号较弱，无法和在室外空旷环境一样用卫星定位的方式获得整个实景地图中各个图像的准确坐标信息。因此，在生成实景地图时，相关技术大多采用室内定位技术获取实景地图的坐标信息。然而，室内定位技术需要提前布置室内定位设备，且在生成实景地图的过程中对现场要求较高，大多数情况下需要清场，防止对信号产生干扰造成实景地图不准确的问题。因此，相关技术在生成实景地图时需要消耗大量人力和时间，实景地图生成效率较低。
因此，如何解决相关技术需要消耗大量人力和时间，实景地图生成效率较低的问题，是本领域技术人员需要解决的技术问题。
发明内容
有鉴于此，本发明的目的在于提供一种实景地图生成方法、实景地图生成装置、实景地图生成设备及计算机可读存储介质，解决了相关技术需要消耗大量人力和时间，实景地图生成效率较低的问题。
为解决上述技术问题，本发明提供了一种实景地图生成方法，包括：
获取实景图像集以及与所述实景图像集对应的锚定点坐标；
利用所述锚定点坐标计算所述实景图像集对应的预设坐标集；
根据所述实景图像集对所述预设坐标集进行运动恢复结构处理，得到目标坐标集；
利用所述实景图像集与所述目标坐标集生成实景地图。
可选地，所述获取实景图像集以及与所述实景图像集对应的锚定点坐标，包括：
获取多张实景图像，并利用所述实景图像组成所述实景图像集；
在所述实景图像集中确定目标实景图像；
获取各个所述目标实景图像对应的图像坐标，并将所述图像坐标确定为所述锚定点坐标。
可选地，所述获取实景图像集以及与所述实景图像集对应的锚定点坐标，包括：
获取实景视频；
按照预设抽样频率从所述实景视频中抽取多个图像帧，并利用所述图像帧组成所述实景图像集；
在所述实景图像集中确定目标图像帧；
获取各个所述目标图像帧对应的图像坐标，并将所述图像坐标确定为所述锚定点坐标。
可选地，所述利用所述锚定点坐标计算所述实景图像集对应的预设坐标集，包括：
确定所述实景图像集对应的坐标计算规则；
利用所述锚定点坐标，根据所述坐标计算规则，计算所述实景图像集中各个图像对应的预设坐标；
利用各个所述预设坐标构建所述预设坐标集。
可选地，所述根据所述实景图像集对所述预设坐标集进行运动恢复结构处理，得到目标坐标集，包括：
对所述实景图像集中的图像进行特征点提取，并对所述特征点进行匹配，得到多个特征点对；
利用所述特征点对进行空间结构恢复操作，得到空间结构；
根据所述空间结构对所述预设坐标集进行坐标修正，得到所述目标坐标集。
可选地，所述根据所述空间结构对所述预设坐标集进行坐标修正，得 到所述目标坐标集，包括：
根据所述空间结构对所述预设坐标集进行坐标修正，得到中间坐标集；
对所述中间坐标集进行最小化误差处理，得到所述目标坐标集。
可选地，所述利用所述实景图像集与所述目标坐标集生成实景地图，包括：
将所述实景图像集中的图像进行拼合处理，得到初始实景地图；
利用所述目标坐标集对所述初始实景地图进行标记，得到所述实景地图。
本发明还提供了一种实景地图生成装置，包括：
获取模块，用于获取实景图像集以及与所述实景图像集对应的锚定点坐标；
计算模块，用于利用所述锚定点坐标计算所述实景图像集对应的预设坐标集；
处理模块，用于根据所述实景图像集对所述预设坐标集进行运动恢复结构处理，得到目标坐标集；
生成模块，用于利用所述实景图像集与所述目标坐标集生成实景地图。
本发明还提供了一种实景地图生成设备，包括存储器和处理器，其中：
所述存储器，用于保存计算机程序；
所述处理器，用于执行所述计算机程序，以实现上述的实景地图生成方法。
本发明还提供了一种计算机可读存储介质，用于保存计算机程序，其中，所述计算机程序被处理器执行时实现上述的实景地图生成方法。
本发明提供的实景地图生成方法，获取实景图像集以及与实景图像集对应的锚定点坐标；利用锚定点坐标计算实景图像集对应的预设坐标集；根据实景图像集对预设坐标集进行运动恢复结构处理，得到目标坐标集；利用实景图像集与目标坐标集生成实景地图。
可见，该方法在获取实景图像集时同时获取对应的锚定点坐标，通过锚定点坐标可以计算得到实景图像集对应的预设坐标集，预设坐标集包括 了实景图像集中各个图像的预设坐标，即大概坐标。对预设坐标集进行运动恢复结构处理，可以根据实景图像集中的实景图像对预设坐标集进行调整，得到实景图像集对应的准确的坐标集，即目标坐标集。基于实景图像集的运动恢复结构处理可以保证目标坐标集的准确性，利用目标坐标集与实景图像集生成实景地图，可以在保证实景地图准确性的基础上，节省了设置室内定位设备以及利用室内定位设备生成地图所需的大量人力和时间，提高了实景地图的生成效率，解决了相关技术需要消耗大量人力和时间，实景地图生成效率较低的问题。
此外，本发明还提供了一种实景地图生成装置、实景地图生成设备及计算机可读存储介质，同样具有上述有益效果。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据提供的附图获得其他的附图。
图1为本发明实施例提供的一种实景地图生成方法流程图；
图2为本发明实施例提供的一种实景图像集和锚定点坐标获取方法流程图；
图3为本发明实施例提供的另一种实景图像集和锚定点坐标获取方法流程图；
图4为本发明实施例提供的一种具体的实景地图生成方法流程图；
图5为本发明实施例提供的一种实景地图生成装置的结构示意图；
图6为本发明实施例提供的一种实景地图生成设备的结构示意图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实 施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
实景图像为记录现实环境的图像，而具有地理位置信息的实景图像，在室外环境中，常被称为街景地图，在室内环境中，常被称为实景地图。大多数的实景地图搭建方案大多数是关注室外环境的，即用于生成街景图像，例如通过车载设备或者无人机采集实景照片，而实景照片对应的地理位置信息一般通过搭载车载设备的车辆或无人机的GPS等卫星定位设备获取。
由于在室内环境中，GPS等卫星定位设备的信号会被削弱，因此会导致定位不准确的问题。为了保证定位的准确性，进而保证实景地图的准确程度，相关技术一般采用室内定位技术生成实景地图。具体的，在需要绘制实景地图的室内放置定位终端，定位终端对应的坐标为准确的地理坐标。利用携带有定位装置的图像获取设备获取图像，在获取图像的过程中，定位终端利用自身的地理坐标以及与图像获取设备之间的位置关系，对图像获取设备进行定位，得到获取到的图像对应的准确地理位置信息，完成实景地图的生成。
然而，使用室内定位技术生成实景地图具有较大的弊端。该方法对生成实景地图的设备具有一定要求，同时，需要在生成地图前放置定位终端，在生成过后回收定位终端；且在生成实景地图的过程中，为了避免人员在定位终端以及图像获取设备之间的走动对设备与终端间的信号强度产生干扰，进而因信号强度变化引起定位偏差等问题，需要对整个空间进行清场，整个实景地图的生成过程需要耗费大量的人力和时间。
为了解决上述问题，本申请提供了一种实景地图生成方法，以及对应的装置、设备和计算机可读存储介质。该方法不需要设置室内定位终端，通过对实景图像集对应的坐标集进行运动恢复结构处理，即可基于实景图像集中的图像对坐标进行修正，保证实景图像集对应的目标坐标集的准确。在保证实景地图准确性的基础上减少了人力和时间的消耗，大大提高了效率。
具体的，在一种可能的实施方式中，请参考图1。图1为本发明实施例提供的一种实景地图生成方法流程图，该方法包括：
S101：获取实景图像集以及与实景图像集对应的锚定点坐标。
实景图像集包括多张实景图像，实景图像用于生成实景地图，实景图像的具体数量本实施例不做限定。需要说明的是，实景图像为连续的图像，即相邻图像之间具有重叠的部分，以便后续对其进行运动恢复结构处理。实景图像集的获取方法本实施例不做限定，例如可以设置有预设路径，按照预设路径获取实景图像，进而组成实景图像集，在可能的实施方式中，可以在获取实景图像的过程中根据实际情况适当偏离预设路径；或者可以不设定固定的预设路径，根据实际需要选择实际路径获取实景图像，进而组成实景图像集。
锚定点坐标与实景图像集相对应，其具体数量为多个，锚定点坐标为实景图像集中的目标实景图像对应的坐标，需要说明的是，锚定点坐标为利用非室内定位技术获取的准确坐标，具体获取方法本实施例不做限定，例如可以利用卫星定位技术获取锚定点坐标；或者可以利用手机或无线通讯信号基站对锚定点进行定位，得到锚定点坐标。锚定点坐标对应的锚定点可以为任意的位置点，例如可以为预设路径的起点、终点、转折点或其他任意点。
由于锚定点坐标为目标实景图像的坐标，因此锚定点坐标对应的锚定点一定处于实景图像集的实际路径上。当采用预设路径的实景图像集获取方法时，本实施例并不限定预设路径和锚定点的确定顺序。具体的，可以先确定预设路径，并在预设路径中选择具有准确坐标的点作为锚定点；或者可以先确定多个锚定点，并基于锚定点设置预设路径。同理，在根据实际需要选择实际路径获取实景图像集时，可以先获取实景图像集，得到实际路径，然后在实际路径上选择多个具有准确坐标的点作为锚定点；或者可以先确定多个锚定点，并在锚定点获取目标实景图像，并利用目标实景图像和其他的实景图像组成实景图像集。
S102：利用锚定点坐标计算实景图像集对应的预设坐标集。
在得到锚定点坐标以及实景图像集后，利用锚定点坐标计算实景图像 集对应的预设坐标集。预设坐标集内包括多个预设坐标，每个预设坐标与实景图像集中的每张实景图像相对应。
预设坐标为实景图像的大概坐标，其与实景图像相对应，具体计算方法本实施例不做限定，例如当存在预设路径时，可以将预设路径中各个锚定点间的长度平均分段，得到多个预设点，并利用锚定点坐标对预设点的坐标进行计算，得到预设坐标，进而组成预设坐标集；在不存在预设路径时，可以利用直线或曲线将各个锚定点相连，得到模拟路径，在模拟路径上确定预设点并计算预设坐标，最终得到预设坐标集。
S103：根据实景图像集对预设坐标集进行运动恢复结构处理，得到目标坐标集。
预设坐标集中的坐标为大概的坐标，其是不准确的。为了生成准确地实景地图，需要对其进行调整，得到目标坐标集，目标坐标集即为实景图像集对应的准确坐标集。具体的，运动恢复结构即为Structure From Motion，SFM，通过在不同图片中定位多个共同特征点，并根据光线直线传播的原则，结合特征点在至少两张图片上像素位置的变化以及图像获取设备的参数，反算出图像获取设备的位置，即实景图像对应的目标位置。设备的参数可以包括但不限于焦距、感光元件尺寸、相机矩阵等。
需要说明的是，运动恢复结构处理可以采用SFM算法，或者可以采用其他类似的算法或程序对预设坐标集进行处理，只需得到目标坐标集即可。本实施例并不限定目标坐标集中目标坐标的表示形式，其与预设坐标集的表示形式相同，即与锚定点坐标的表示形式相同，例如可以为经纬度坐标形式；或者可以根据实际需要选择其他形式，例如可以为自定义坐标系中的坐标表示形式。
S104：利用实景图像集与目标坐标集生成实景地图。
在得到目标坐标集后，可以利用实景图像集与目标坐标集生成实景地图，本实施例并不限定实景地图的具体生成方法，可以参考相关技术，在此不再赘述。本实施例中不需要利用室内定位技术获取准确的位置信息，而是利用运动恢复结构处理得到目标坐标集，在保证了实景地图准确性的同时，节省了设置室内定位设备以及利用室内定位设备生成地图所需的大 量人力和时间。
应用本发明实施例提供的实景地图生成方法，在获取实景图像集时同时获取对应的锚定点坐标，通过锚定点坐标可以计算得到实景图像集对应的预设坐标集，预设坐标集包括了实景图像集中各个图像的预设坐标，即大概坐标。对预设坐标集进行运动恢复结构处理，可以根据实景图像集中的图像对预设坐标集进行调整，得到实景图像集对应的准确的坐标集，即目标坐标集。基于实景图像集的运动恢复结构处理可以保证目标坐标集的准确性，利用目标坐标集与实景图像集生成实景地图，可以在保证实景地图准确性的基础上，节省了设置室内定位设备以及利用室内定位设备生成地图所需的大量人力和时间，提高了实景地图的生成效率，解决了相关技术需要消耗大量人力和时间，实景地图生成效率较低的问题。
基于上述实施例，在一种可能的实施方式中，可以采用拍照的方式获取实景图像集。具体请参考图2，图2为本发明实施例提供的一种实景图像集和锚定点坐标获取方法流程图，包括：
S201：获取多张实景图像，并利用实景图像组成实景图像集。
在本实施例中，通过拍照获取实景图像的方法组成实景图像集，本实施例并不限定获取实景图像的具体途径，例如可以将图像获取设备设置于机器人上，由机器人利用图像获取设备获取实景图像；或者可以利用人工的方式获取实景图像。进一步，为了保证实景图像的清晰，可以利用具有防抖功能的图像获取设备获取实景图像；或者可以利用安装有云台的图像获取设备获取实景图像，本实施例并不对图像获取设备的具体内容做出限定，例如可以为全景相机、普通相机或手机等。
本实施例并不限定获取实景图像的具体方法，例如可以在运动中按照预设时间间隔获取实景图像，运动可以为匀速运动或非匀速运动；或者可以按照预设距离间隔获取实景图像。在获取到实景图像后，利用其组成实景图像集。
S202：在实景图像集中确定目标实景图像。
需要说明的是，本实施例中目标实景图像为具有准确坐标的实景图像， 即在锚定点拍摄的实景图像。因此在获取实景图像集后，在其中确定目标实景图像，以便获取锚定点坐标。
S203：获取各个目标实景图像对应的图像坐标，并将图像坐标确定为锚定点坐标。
在确定目标实景图像后，获取其对应的图像坐标，则该图像坐标即为锚定点坐标。需要说明的是，本实施例并不限定锚定点与目标实景图像的确定顺序，即可以先确定锚定点，并将锚定点处拍摄的实景图像确定为目标实景图像；或者可以先确定目标实景图像，即将全部或部分具有准确坐标的实景图像确定为目标实景图像，并将其拍摄位置确定为锚定点，将其坐标确定为锚定点坐标。
在另一种可能的实施方式中，为了提高目标实景图像的选择灵活度和实景图像集的选择灵活度，可以采用抽取图像帧的方式获取实景图像集。可以参考图3，图3为本发明实施例提供的另一种实景图像集和锚定点坐标获取方法流程图，包括：
S301：获取实景视频。
在本实施例中，通过在实景视频中抽取图像帧的方式获取实景图像集，实景视频的获取方法本实施例不做限定，可以参考上述实景图像的获取方法。
S302：按照预设抽样频率从实景视频中抽取多个图像帧，并利用图像帧组成实景图像集。
具体的，可以按照预设抽样频率从实景视频中抽取多个图像帧，进而组成实景图像集。预设抽样频率的具体大小可以根据实际情况设定，本实施例不做限定。由于实景视频中具有多个图像帧，因此可以提高实景图像集的选择灵活度。需要说明的是，图像帧只是实景图像的一种特殊形式，图像帧本身仍是实景图像。
在另一种可能的实施方式中，可以根据实际需要设定抽样规则，按照抽样规则从实景视频中抽取图像帧，或者可以由人工选择图像帧。在抽取图像帧后，还可以对实景图像集中的图像帧进行新增、删除或替换，本实施例对此不做限定。
S303：在实景图像集中确定目标图像帧。
在得到实景图像集后，可以从中确定目标图像帧，目标图像帧即为在锚定点获取的图像帧。由于图像帧的选择更具灵活性，因此目标图像帧的选择也可以更加灵活。
S304：获取各个目标图像帧对应的图像坐标，并将图像坐标确定为锚定点坐标。
本实施例中，目标图像帧即为目标实景图像，目标图像帧的坐标即为锚定点坐标。
基于上述实施例，在一种可能的实施方式中，在得到锚定点坐标后，可以设置有坐标计算规则，利用坐标计算规则计算预设坐标。具体请参考图4，图4为本发明实施例提供的一种具体的实景地图生成方法流程图，包括：
S401：确定实景图像集对应的坐标计算规则。
坐标计算规则用于计算预设坐标，在本实施例中，坐标计算规则与实景图像集相对应，本实施例并不限定坐标计算规则的具体内容，具体的，可以根据实景图像集的获取方式确定坐标计算规则。例如在具有预设路径时，可以按照预设路径设置坐标计算规则，在某些可能的实施方式中，获取实景图像时适当偏离了预设路径，在这种情况下可以仍然按照预设路径设置坐标计算规则，或者可以根据预设路径和偏移量设置坐标计算规则。当不存在预设路径时，可以利用直线或曲线将锚定点相连，得到模拟路径，以便根据模拟路径设置坐标计算规则。此外，还可以利用预设获取间隔确定各个实景图像的分布，以便计算各个预设坐标。预设获取间隔可以为时间间隔，例如一秒；或者可以为距离间隔，例如50厘米。
S402：利用锚定点坐标，根据坐标计算规则，计算实景图像集中各个实景图像对应的预设坐标。
利用锚定点坐标根据坐标计算规则对预设坐标进行计算，即可得到实景图像集中各个实景图像对应的预设坐标。
S403：利用各个预设坐标构建预设坐标集。
在得到预设坐标后，利用其与锚定点坐标一起构建预设坐标集。
在本实施例中，运动恢复结构处理的过程包括S404、S405和S406三个步骤，具体的：
S404：对实景图像集中的实景图像进行特征点提取，并对特征点进行匹配，得到多个特征点对。
在对预设坐标集进行修正前，需要对实景图像集中的实景图像进行特征点提取，并将不同实景图像中的特征点进行匹配，相同的特征点即可构成特征点对，以便在后续构建空间视觉结构，即稀疏点云。需要说明的是，由于特征点对中的特征点为不同实景图像中的同一特征，因此这多张实景图像为相邻的实景图像。
S405：利用特征点对进行空间结构恢复操作，得到空间结构。
空间结构即为实景图像所记录的空间结构，也可以被称为稀疏点云。在得到特征点对后，利用其进行空间结构恢复操作，可以确定各个特征点之间的相对空间位置，而该相对空间位置即为特征点对应的空间结构。
S406：根据空间结构对预设坐标集进行坐标修正，得到目标坐标集。
在得到空间结构后，利用其对预设标记中的部分或全部预设坐标进行修正，即可得到目标坐标集。具体的，坐标修正即利用空间结构进行基于预设坐标集的坐标推算，得到各个实景图像对应的拍摄位置坐标，或称为图像坐标，该拍摄位置坐标即为目标坐标集中的目标坐标。
在一种可能的实施方式中，还可以进行最小化误差处理，以便进一步提高目标坐标集的准确性，此时，S406步骤可以包括：
S4061：根据空间结构对预设坐标集进行坐标修正，得到中间坐标集。
在对预设坐标集进行坐标修正后，可以得到一个或多个中间坐标集，不同的中间坐标集可以为对预设坐标集进行不同的坐标修正得到的坐标集，其数量和具体内容本实施例不做限定。
S4062：对中间坐标集进行最小化误差处理，得到目标坐标集。
在得到中间坐标集后，对其进行最小化误差处理，以便减少误差，最终得到目标坐标集。最小化误差处理也可以称为全局最优策略，具体的处理过程本实施例不做限定，可以参考相关技术。
S407：将实景图像集中的实景图像进行拼合处理，得到初始实景地图。
实景地图一般为整张图像，因此在得到目标坐标集后，为了生成实景地图，需要将实景图像集中的实景图像进行拼合处理，以便得到整张的实景地图，即初始实景地图。
S408：利用目标坐标集对初始实景地图进行标记，得到实景地图。
通过利用目标坐标集对初始实景地图进行标记，可以使其附带有准确的坐标信息，最终得到实景地图。
下面对本发明实施例提供的实景地图生成装置进行介绍，下文描述的实景地图生成装置与上文描述的实景地图生成方法可相互对应参照。
请参考图5，图5为本发明实施例提供的一种实景地图生成装置的结构示意图，包括：
获取模块510，用于获取实景图像集以及与实景图像集对应的锚定点坐标；
计算模块520，用于利用锚定点坐标计算实景图像集对应的预设坐标集；
处理模块530，用于根据实景图像集对预设坐标集进行运动恢复结构处理，得到目标坐标集；
生成模块540，用于利用实景图像集与目标坐标集生成实景地图。
可选地，获取模块510，包括：
第一图像集获取单元，用于获取多张实景图像，并利用实景图像组成实景图像集；
第一确定单元，用于在实景图像集中确定目标实景图像；
第一坐标获取单元，用于获取各个目标实景图像对应的图像坐标，并将图像坐标确定为锚定点坐标。
可选地，获取模块510，包括：
视频获取单元，用于获取实景视频；
第二图像集获取单元，用于按照预设抽样频率从实景视频中抽取多个图像帧，并利用图像帧组成实景图像集；
第二确定单元，用于在实景图像集中确定目标图像帧；
第二坐标获取单元，用于获取各个目标图像帧对应的图像坐标，并将图像坐标确定为锚定点坐标。
可选地，计算模块520，包括：
计算规则确定单元，用于确定实景图像集对应的坐标计算规则；
预设坐标计算单元，用于利用锚定点坐标，根据坐标计算规则，计算实景图像集中各个图像对应的预设坐标；
预设坐标集构建单元，用于利用各个预设坐标构建预设坐标集。
可选地，处理模块530，包括：
特征点提取单元，用于对实景图像集中的图像进行特征点提取，并对特征点进行匹配，得到多个特征点对；
空间结构生成单元，用于利用特征点对进行空间结构恢复操作，得到空间结构；
坐标修正单元，用于根据空间结构对预设坐标集进行坐标修正，得到目标坐标集。
可选地，坐标修正单元，包括：
修正子单元，用于根据空间结构对预设坐标集进行坐标修正，得到中间坐标集；
误差处理子单元，用于对中间坐标集进行最小化误差处理，得到目标坐标集。
可选地，生成模块540，包括：
拼合单元，用于将实景图像集中的图像进行拼合处理，得到初始实景地图；
标记单元，用于利用目标坐标集对初始实景地图进行标记，得到实景地图。
应用本发明实施例提供的实景地图生成装置，在获取实景图像集时同时获取对应的锚定点坐标，通过锚定点坐标可以计算得到实景图像集对应的预设坐标集，预设坐标集包括了实景图像集中各个图像的预设坐标，即大概坐标。对预设坐标集进行运动恢复结构处理，可以根据实景图像集中 的图像对预设坐标集进行调整，得到实景图像集对应的准确的坐标集，即目标坐标集。基于实景图像集的运动恢复结构处理可以保证目标坐标集的准确性，利用目标坐标集与实景图像集生成实景地图，可以在保证实景地图准确性的基础上，节省了设置室内定位设备以及利用室内定位设备生成地图所需的大量人力和时间，提高了实景地图的生成效率，解决了相关技术需要消耗大量人力和时间，实景地图生成效率较低的问题。
下面对本发明实施例提供的实景地图生成设备进行介绍，下文描述的实景地图生成设备与上文描述的实景地图生成方法可相互对应参照。
请参考图6，图6为本发明实施例提供的一种实景地图生成设备的结构示意图。其中实景地图生成设备600可以包括处理器601和存储器602，还可以进一步包括多媒体组件603、信息输入/信息输出(I/O)接口604以及通信组件605中的一种或多种。
其中，处理器601用于控制实景地图生成设备600的整体操作，以完成上述的实景地图生成方法中的全部或部分步骤；存储器602用于存储各种类型的数据以支持在实景地图生成设备600的操作，这些数据例如可以包括用于在该实景地图生成设备600上操作的任何应用程序或方法的指令，以及应用程序相关的数据。该存储器602可以由任何类型的易失性或非易失性存储设备或者它们的组合实现，例如静态随机存取存储器(Static Random Access Memory，SRAM)、电可擦除可编程只读存储器(Electrically Erasable Programmable Read-Only Memory，EEPROM)、可擦除可编程只读存储器(Erasable Programmable Read-Only Memory，EPROM)、可编程只读存储器(Programmable Read-Only Memory，PROM)、只读存储器(Read-Only Memory，ROM)、磁存储器、快闪存储器、磁盘或光盘中的一种或多种。
多媒体组件603可以包括屏幕和音频组件。其中屏幕例如可以是触摸屏，音频组件用于输出和/或输入音频信号。例如，音频组件可以包括一个麦克风，麦克风用于接收外部音频信号。所接收的音频信号可以被进一步存储在存储器602或通过通信组件605发送。音频组件还包括至少一个扬声器，用于输出音频信号。I/O接口604为处理器601和其他接口模块之间提 供接口，上述其他接口模块可以是键盘，鼠标，按钮等。这些按钮可以是虚拟按钮或者实体按钮。通信组件605用于实景地图生成设备600与其他设备之间进行有线或无线通信。无线通信，例如Wi-Fi，蓝牙，近场通信(Near Field Communication，简称NFC)，2G、3G或4G，或它们中的一种或几种的组合，因此相应的该通信组件605可以包括：Wi-Fi部件，蓝牙部件，NFC部件。
实景地图生成设备600可以被一个或多个应用专用集成电路(Application Specific Integrated Circuit，简称ASIC)、数字信号处理器(Digital Signal Processor，简称DSP)、数字信号处理设备(Digital Signal Processing Device，简称DSPD)、可编程逻辑器件(Programmable Logic Device，简称PLD)、现场可编程门阵列(Field Programmable Gate Array，简称FPGA)、控制器、微控制器、微处理器或其他电子元件实现，用于执行上述实施例给出的实景地图生成方法。
下面对本发明实施例提供的计算机可读存储介质进行介绍，下文描述的计算机可读存储介质与上文描述的实景地图生成方法可相互对应参照。
本发明还提供一种计算机可读存储介质，计算机可读存储介质上存储有计算机程序，计算机程序被处理器执行时实现上述的实景地图生成方法的步骤。
该计算机可读存储介质可以包括：U盘、移动硬盘、只读存储器(Read-Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、磁碟或者光盘等各种可以存储程序代码的介质。
本说明书中各个实施例采用递进的方式描述，每个实施例重点说明的都是与其它实施例的不同之处，各个实施例之间相同或相似部分互相参见即可。对于实施例公开的装置而言，由于其与实施例公开的方法相对应，所以描述的比较简单，相关之处参见方法部分说明即可。
专业人员还可以进一步意识到，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、计算机软件或者二者的结合来 实现，为了清楚地说明硬件和软件的可互换性，在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件的方式来执行，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，但是这种实现不应该认为超出本发明的范围。
结合本文中所公开的实施例描述的方法或算法的步骤可以直接用硬件、处理器执行的软件模块，或者二者的结合来实施。软件模块可以置于随机存储器(RAM)、内存、只读存储器(ROM)、电可编程ROM、电可擦除可编程ROM、寄存器、硬盘、可移动磁盘、CD-ROM、或技术领域内所公知的任意其它形式的存储介质中。
最后，还需要说明的是，在本文中，诸如第一和第二等之类的关系属于仅仅用来将一个实体或者操作与另一个实体或者操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且，术语包括、包含或者其他任何变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者设备所固有的要素。
以上对本发明所提供的实景地图生成方法、实景地图生成装置、实景地图生成设备和计算机可读存储介质进行了详细介绍，本文中应用了具体个例对本发明的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本发明的方法及其核心思想；同时，对于本领域的一般技术人员，依据本发明的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本发明的限制。

Claims (10)

  1. 一种实景地图生成方法，其特征在于，包括：
    获取实景图像集以及与所述实景图像集对应的锚定点坐标；
    利用所述锚定点坐标计算所述实景图像集对应的预设坐标集；
    根据所述实景图像集对所述预设坐标集进行运动恢复结构处理，得到目标坐标集；
    利用所述实景图像集与所述目标坐标集生成实景地图。
  2. 根据权利要求1所述的实景地图生成方法，其特征在于，所述获取实景图像集以及与所述实景图像集对应的锚定点坐标，包括：
    获取多张实景图像，并利用所述实景图像组成所述实景图像集；
    在所述实景图像集中确定目标实景图像；
    获取各个所述目标实景图像对应的图像坐标，并将所述图像坐标确定为所述锚定点坐标。
  3. 根据权利要求1所述的实景地图生成方法，其特征在于，所述获取实景图像集以及与所述实景图像集对应的锚定点坐标，包括：
    获取实景视频；
    按照预设抽样频率从所述实景视频中抽取多个图像帧，并利用所述图像帧组成所述实景图像集；
    在所述实景图像集中确定目标图像帧；
    获取各个所述目标图像帧对应的图像坐标，并将所述图像坐标确定为所述锚定点坐标。
  4. 根据权利要求1所述的实景地图生成方法，其特征在于，所述利用所述锚定点坐标计算所述实景图像集对应的预设坐标集，包括：
    确定所述实景图像集对应的坐标计算规则；
    利用所述锚定点坐标，根据所述坐标计算规则，计算所述实景图像集中各个实景图像对应的预设坐标；
    利用各个所述预设坐标构建所述预设坐标集。
  5. 根据权利要求1所述的实景地图生成方法，其特征在于，所述根据所述实景图像集对所述预设坐标集进行运动恢复结构处理，得到目标坐标 集，包括：
    对所述实景图像集中的实景图像进行特征点提取，并对所述特征点进行匹配，得到多个特征点对；
    利用所述特征点对进行空间结构恢复操作，得到空间结构；
    根据所述空间结构对所述预设坐标集进行坐标修正，得到所述目标坐标集。
  6. 根据权利要求5所述的实景地图生成方法，其特征在于，所述根据所述空间结构对所述预设坐标集进行坐标修正，得到所述目标坐标集，包括：
    根据所述空间结构对所述预设坐标集进行坐标修正，得到中间坐标集；
    对所述中间坐标集进行最小化误差处理，得到所述目标坐标集。
  7. 根据权利要求1所述的实景地图生成方法，其特征在于，所述利用所述实景图像集与所述目标坐标集生成实景地图，包括：
    将所述实景图像集中的实景图像进行拼合处理，得到初始实景地图；
    利用所述目标坐标集对所述初始实景地图进行标记，得到所述实景地图。
  8. 一种实景地图生成装置，其特征在于，包括：
    获取模块，用于获取实景图像集以及与所述实景图像集对应的锚定点坐标；
    计算模块，用于利用所述锚定点坐标计算所述实景图像集对应的预设坐标集；
    处理模块，用于根据所述实景图像集对所述预设坐标集进行运动恢复结构处理，得到目标坐标集；
    生成模块，用于利用所述实景图像集与所述目标坐标集生成实景地图。
  9. 一种实景地图生成设备，其特征在于，包括存储器和处理器，其中：
    所述存储器，用于保存计算机程序；
    所述处理器，用于执行所述计算机程序，以实现如权利要求1至7任一项所述的实景地图生成方法。
  10. 一种计算机可读存储介质，其特征在于，用于保存计算机程序， 其中，所述计算机程序被处理器执行时实现如权利要求1至7任一项所述的实景地图生成方法。
PCT/CN2020/089553 2020-05-11 2020-05-11 一种实景地图生成方法、装置、设备及可读存储介质 WO2021226780A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202080001086.3A CN111801664A (zh) 2020-05-11 2020-05-11 一种实景地图生成方法、装置、设备及可读存储介质
PCT/CN2020/089553 WO2021226780A1 (zh) 2020-05-11 2020-05-11 一种实景地图生成方法、装置、设备及可读存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2020/089553 WO2021226780A1 (zh) 2020-05-11 2020-05-11 一种实景地图生成方法、装置、设备及可读存储介质

Publications (1)

Publication Number Publication Date
WO2021226780A1 true WO2021226780A1 (zh) 2021-11-18

Family

ID=72834259

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/089553 WO2021226780A1 (zh) 2020-05-11 2020-05-11 一种实景地图生成方法、装置、设备及可读存储介质

Country Status (2)

Country Link
CN (1) CN111801664A (zh)
WO (1) WO2021226780A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108205319A (zh) * 2016-12-19 2018-06-26 三星电子株式会社 可移动对象及其控制方法
US20180261005A1 (en) * 2017-03-07 2018-09-13 Baidu Online Network Technology (Beijing) Co., Ltd. Method and Apparatus for Constructing Three-Dimensional Map
CN110260857A (zh) * 2019-07-02 2019-09-20 北京百度网讯科技有限公司 视觉地图的校准方法、装置及存储介质
CN110738143A (zh) * 2019-09-27 2020-01-31 Oppo广东移动通信有限公司 定位方法及装置、设备、存储介质

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106646562A (zh) * 2016-09-09 2017-05-10 华东师范大学 高精度三维实景室内外一体化定位方法及装置
JP2019207467A (ja) * 2018-05-28 2019-12-05 株式会社イームズラボ ３次元マップ補正装置、３次元マップ補正方法及び３次元マップ補正プログラム

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108205319A (zh) * 2016-12-19 2018-06-26 三星电子株式会社 可移动对象及其控制方法
US20180261005A1 (en) * 2017-03-07 2018-09-13 Baidu Online Network Technology (Beijing) Co., Ltd. Method and Apparatus for Constructing Three-Dimensional Map
CN110260857A (zh) * 2019-07-02 2019-09-20 北京百度网讯科技有限公司 视觉地图的校准方法、装置及存储介质
CN110738143A (zh) * 2019-09-27 2020-01-31 Oppo广东移动通信有限公司 定位方法及装置、设备、存储介质

Also Published As

Publication number Publication date
CN111801664A (zh) 2020-10-20

Similar Documents

Publication Publication Date Title
US10495756B2 (en) Multi-camera laser scanner
US20160178728A1 (en) Indoor Positioning Terminal, Network, System and Method
JP5056469B2 (ja) 画像管理装置
US8687070B2 (en) Image capture device having tilt and/or perspective correction
JP5736526B2 (ja) 電子マップに基づく位置検索方法及び装置
CN111436208B (zh) 一种测绘采样点的规划方法、装置、控制终端及存储介质
CN111627114A (zh) 室内视觉导航方法、装置、系统及电子设备
US20190068877A1 (en) Mobile terminal image synthesis method, mobile terminal image synthesis apparatus and mobile terminal
TW201621701A (zh) 拍照方法及智慧終端機、雲端伺服器
CN109520500A (zh) 一种基于终端拍摄图像匹配的精确定位及街景库采集方法
WO2017133147A1 (zh) 一种实景地图的生成方法、推送方法及其装置
CN103843329A (zh) 用于立体图像对的有条件显示的方法和设备
US9749597B2 (en) Precise target positioning in geographical imaging
CN103581562A (zh) 全景拍摄的方法和装置
JP5812509B2 (ja) 地図表示装置、地図表示方法及びプログラム
KR20100060472A (ko) 위치 측정 방법 및 이를 이용한 휴대 단말기
WO2021226780A1 (zh) 一种实景地图生成方法、装置、设备及可读存储介质
CN112839165B (zh) 人脸跟踪摄像的实现方法、装置、计算机设备和存储介质
WO2024012405A1 (zh) 一种标定方法及装置
CN104954679A (zh) 一种图片拍摄的控制方法、装置及终端
JP2005106953A (ja) 電子地図作成装置、電子地図作成方法、電子地図作成プログラム、および該プログラムを記録した記録媒体
WO2020235063A1 (ja) 画像処理方法、画像処理装置およびプログラム
CN107655458B (zh) 基于gis的全景图场景自动关联方法
CN114339179B (zh) 投影校正方法、装置、存储介质以及投影设备
CN116228860A (zh) 目标物的地理位置预测方法、装置、设备及存储介质

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: 20935025

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: 20935025

Country of ref document: EP

Kind code of ref document: A1