KR102616085B1 - Apparatus and method for deciding validity of camera pose using visual localization - Google Patents

Abstract

According to some embodiments of the present disclosure, there is provided a method for visual localization performed by a computing device including a processor, the method comprising: obtaining a first query image corresponding to a first time from a first device; Obtaining first comparison data by comparing first query feature point information of the first query image obtained using a pre-learned feature point acquisition model with reference feature point information of a pre-stored first reference image; estimating a first camera pose of the first device based on the first comparison data; Obtaining a second query image corresponding to a second time different from the first time from the first device; Obtaining second comparison data by comparing second query feature point information of the second query image obtained using the pre-learned feature point acquisition model with the first query feature point information; Estimating a second camera pose of the first device related to the second query image based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device; and determining an effective value of at least one of the first camera pose and the second camera pose based on the first comparison data, the second comparison data, and the information of the inertial measurement unit. .

Description

Method and apparatus for determining validity of camera pose using visual localization {APPARATUS AND METHOD FOR DECIDING VALIDITY OF CAMERA POSE USING VISUAL LOCALIZATION}

This disclosure relates to image processing and more specifically to visual localization.

As the demand for Location Based Services increases, the need for accurate location information has increased. The most common way to determine location on mobile devices and mobile platforms is GNSS (Global Navigation Satellite System). However, because GNSS signals may be blocked by obstacles in an indoor environment, there is a limitation that they can only be easily used in an outdoor environment.

Although various technologies have been proposed to perform location recognition indoors, the reality is that many indoor location recognition techniques do not go beyond the fingerprint printing-based location recognition algorithm using wireless signals. In this method, the collected data related to Wi-Fi RSS (Received Signal Strength) or MFS (Magnetic Field Strength) is compared with data from a fingerprinting database. Fingerprinting-based systems have the advantage of being easy to build, but it can be difficult to maintain good performance because the signal pattern itself is affected by changes in the system environment. To overcome the deficiencies of these fingerprint-based systems, many alternatives have been proposed, including Optical, RFID (Radio Frequency Identification), Bluetooth Beacons, ZigBee, and Pseudo Satellite, but these alternatives also have difficulty achieving high accuracy in complex indoor environments. It is evaluated as difficult.

Recently, research on VPS (Visual Positioning System) has been actively conducted as an alternative to implement highly accurate position estimation even in indoor environments. VPS technology can also be expressed as visual localization technology, which refers to a technology that estimates the current location or pose of a device using images taken indoors or outdoors. Pose estimation of a device (or camera) refers to determining the translation and rotation information of a dynamically changing camera viewpoint. This camera pose estimation technology is being used in a variety of fields such as mixed reality, augmented reality, robot navigating, and 3-Dimensional Scene Reconstruction.

Republic of Korea registered patent 10-2225093 (registered on March 3, 2021)

This disclosure has been made in response to the above-described background technology, and is intended to determine the effectiveness of a camera pose using visual localization.

The technical problems of the present disclosure are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the description below.

According to some embodiments of the present disclosure for solving the above-mentioned problems, there is provided a method for visual localization performed by a computing device including a processor, wherein a first signal corresponding to a first time is received from a first device. 1 Obtaining a query image; Obtaining first comparison data by comparing first query feature point information of the first query image obtained using a pre-learned feature point acquisition model with reference feature point information of a pre-stored first reference image; estimating a first camera pose of the first device based on the first comparison data; Obtaining a second query image corresponding to a second time different from the first time from the first device; Obtaining second comparison data by comparing second query feature point information of the second query image obtained using the pre-learned feature point acquisition model with the first query feature point information; Estimating a second camera pose of the first device related to the second query image based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device; and determining an effective value of at least one of the first camera pose and the second camera pose based on the first comparison data, the second comparison data, and the information of the inertial measurement unit. .

Alternatively, the pre-trained feature point acquisition model may be learned in advance through deep learning or machine learning to acquire feature points from a specific image included in specific input data.

Alternatively, acquiring the first comparison data may include matching reference image feature point information of each of a plurality of pre-stored reference images with the first query feature point information; determining a matching reference image that matches the first query feature point information among the plurality of pre-stored reference images as the first reference image; and obtaining the first comparison data by comparing the first query feature point information with reference feature point information of the first reference image.

Alternatively, the step of estimating the first camera pose includes estimating the first camera pose based on 2D matching coordinate values and 3D matching coordinate values of each of the plurality of matching feature points included in the first comparison data. Step; may include.

Alternatively, the 2D matching coordinate values may include two-dimensional pixel coordinate values in the first query image.

Alternatively, the 3D matching coordinate values may include three-dimensional coordinate values in a coordinate system of a first reference device associated with the pre-stored first reference image.

Alternatively, the information of the inertial measurement unit may include information related to movement of the first device from the first time to the second time.

Alternatively, acquiring the second query image may include acquiring the second query image from a second device positioned based on the first camera pose and information of the inertial measurement unit. there is.

Alternatively, determining a valid value of at least one of the first camera pose or the second camera pose may, based on the first comparison data, be associated with the first device and the pre-stored first reference image. estimating a relative camera pose between first reference devices; Obtaining a first query 3D estimated coordinate value of the first query image based on the relative camera pose and a first reference 3D coordinate value of the pre-stored first reference image; Obtaining a second query 3D estimated coordinate value of the second query image based on the first query 3D estimated coordinate value and information of the inertial measurement unit; Obtaining a second query 2D estimated coordinate value of the second query image from the second query 3D estimated coordinate value using a predetermined algorithm; Obtaining second query 2D real coordinate values of the second query image based on the second comparison data; and determining a value of reprojection error calculated by comparing the second query 2D estimated coordinate value and the second query 2D actual coordinate value as a valid value of at least one of the first camera pose and the second camera pose; may include.

Alternatively, if a valid value of at least one of the first camera pose or the second camera pose corresponds to a predetermined condition, re-estimating at least one of the first camera pose or the second camera pose; More may be included.

Alternatively, a computer program stored on a computer-readable storage medium, the computer program comprising instructions for causing a processor of a computing device for visual localization to perform the following steps, said steps being: Obtaining a first query image corresponding to a first time from one device; Obtaining first comparison data by comparing first query feature point information of the first query image obtained using a pre-learned feature point acquisition model with reference feature point information of a pre-stored first reference image; estimating a first camera pose of the first device based on the first comparison data; Obtaining a second query image corresponding to a second time different from the first time from the first device; Obtaining second comparison data by comparing second query feature point information of the second query image obtained using the pre-learned feature point acquisition model with the first query feature point information; Estimating a second camera pose of the first device related to the second query image based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device; and determining an effective value of at least one of the first camera pose and the second camera pose based on the first comparison data, the second comparison data, and the information of the inertial measurement unit. .

Alternatively, a computing device for visual localization comprising: a processor comprising at least one core; a memory storing a computer program executable by the processor; and a network unit, wherein the processor acquires a first query image corresponding to a first time from a first device, and executes a first query of the first query image obtained using a pre-learned feature point acquisition model. Compare feature point information with reference feature point information of a pre-stored first reference image to obtain first comparison data, estimate a first camera pose of the first device based on the first comparison data, and Obtaining a second query image corresponding to a second time different from the first time, and second query feature point information and the first query feature point of the second query image obtained using the pre-learned feature point acquisition model Compare the information to obtain second comparison data, and based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device, the Estimating a second camera pose of a first device, and based on the first comparison data, the second comparison data, and information from the inertial measurement unit, at least one of the first camera pose or the second camera pose Valid values can be determined.

A technique according to an embodiment of the present disclosure can determine the effectiveness of a camera pose using visual localization.

The effects that can be obtained from the present disclosure are not limited to the effects mentioned above, and other effects not mentioned can be clearly understood by those skilled in the art from the description below. .

Various aspects will now be described with reference to the drawings, where like reference numerals are used to collectively refer to like elements. In the following examples, for purposes of explanation, numerous specific details are set forth to provide a comprehensive understanding of one or more aspects. However, it will be clear that such aspect(s) may be practiced without these specific details.
1 schematically illustrates a system for determining the validity of a camera pose using visual localization in accordance with some embodiments of the present disclosure.
Figure 2 is a schematic diagram illustrating network functions according to some embodiments of the present disclosure.
FIG. 3 is a diagram illustrating a process for estimating a camera pose by a computing device according to some embodiments of the present disclosure.
FIG. 4 is a diagram illustrating a method for minimizing reprojection error of a camera pose by a computing device according to some embodiments of the present disclosure.
5 and 6 exemplarily illustrate a method for determining an effective value of a camera pose according to some embodiments of the present disclosure.
Figure 7 depicts a brief, general schematic diagram of an example computing environment in which embodiments of the present disclosure may be implemented.

Various embodiments are now described with reference to the drawings. In this specification, various descriptions are presented to provide an understanding of the disclosure. However, it is clear that these embodiments may be practiced without these specific descriptions.

As used herein, the terms “component,” “module,” “system,” and the like refer to a computer-related entity, hardware, firmware, software, a combination of software and hardware, or an implementation of software. For example, a component may be, but is not limited to, a process running on a processor, a processor, an object, a thread of execution, a program, and/or a computer. For example, both an application running on a computing device and the computing device can be a component. One or more components may reside within a processor and/or thread of execution. A component may be localized within one computer. A component may be distributed between two or more computers. Additionally, these components can execute from various computer-readable media having various data structures stored thereon. Components can transmit signals, for example, with one or more data packets (e.g., data and/or signals from one component interacting with other components in a local system, a distributed system, to other systems and over a network such as the Internet). Depending on the data being transmitted, they may communicate through local and/or remote processes.

Additionally, the term “or” is intended to mean an inclusive “or” and not an exclusive “or.” That is, unless otherwise specified or clear from context, “X utilizes A or B” is intended to mean one of the natural implicit substitutions. That is, either X uses A; X uses B; Or, if X uses both A and B, “X uses A or B” can apply to either of these cases. Additionally, the term “and/or” as used herein should be understood to refer to and include all possible combinations of one or more of the related listed items.

Additionally, the terms “comprise” and/or “comprising” should be understood to mean that the corresponding feature and/or element is present. However, the terms “comprise” and/or “comprising” should be understood as not excluding the presence or addition of one or more other features, elements and/or groups thereof. Additionally, unless otherwise specified or the context is clear to indicate a singular form, the singular terms herein and in the claims should generally be construed to mean “one or more.”

And, the term “at least one of A or B” should be interpreted to mean “if it contains only A,” “if it contains only B,” or “if it is a combination of A and B.”

Those skilled in the art will additionally recognize that the various illustrative logical blocks, components, modules, circuits, means, logic, and algorithm steps described in connection with the embodiments disclosed herein may be implemented using electronic hardware, computer software, or a combination of both. It must be recognized that it can be implemented with To clearly illustrate the interchangeability of hardware and software, various illustrative components, blocks, configurations, means, logics, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented in hardware or software will depend on the specific application and design constraints imposed on the overall system. A skilled technician can implement the described functionality in a variety of ways for each specific application. However, such implementation decisions should not be construed as causing a departure from the scope of the present disclosure.

The description of the presented embodiments is provided to enable anyone skilled in the art to use or practice the present invention. Various modifications to these embodiments will be apparent to those skilled in the art. The general principles defined herein may be applied to other embodiments without departing from the scope of the disclosure. Therefore, the present invention is not limited to the embodiments presented herein. The present invention is to be interpreted in the broadest scope consistent with the principles and novel features presented herein.

The description of the presented embodiments is provided to enable anyone skilled in the art to use or practice the present invention. Various modifications to these embodiments will be apparent to those skilled in the art. The general principles defined herein may be applied to other embodiments without departing from the scope of the disclosure. Therefore, the present invention is not limited to the embodiments presented herein. The present invention is to be interpreted in the broadest scope consistent with the principles and novel features presented herein.

In the present disclosure, terms represented by N, such as first, second, or third, are used to distinguish a plurality of entities. For example, the entities expressed as first and second may be the same or different from each other. Terms expressed as 1-1, 1-2, and 2-1, 2-2 may also be used to distinguish a plurality of entities.

In the present disclosure, visual localization refers to a technology for estimating the current location or pose of a device using images taken indoors or outdoors. Pose estimation for a user device (or camera) refers to determining translation and rotation information of a dynamically changing camera viewpoint. As an exemplary implementation method rather than a limitation on visual localization, a Visual Positioning System (VPS) can be used.

The query image in the query data in the present disclosure includes an image captured by a device, and a reference image stored in a computing device (e.g., server, etc.) and a device corresponding to the query image using the query image. The pose of can be determined or estimated.

A reference image in the present disclosure may include a pre-stored image used for pose estimation of the device that captured the query image. The reference image may include actual images taken in a specific area to implement, for example, augmented reality or virtual reality. These reference images may be stored on the computing device. Metadata mapped to the reference image can also be used to estimate the pose of the user device.

Metadata that may be included in query data in the present disclosure may mean additional information used in performing visual localization. For example, metadata may include additional information other than the image, such as information related to the device, pixel coordinate information of feature points, descriptor of feature points, landmark information obtained from the image, character information obtained from the image, etc. . For example, by utilizing metadata of the query image and/or reference image, the candidate group of reference images to be matched with the query image may be reduced. As another example, by utilizing metadata of the query image and/or reference image, a more suitable camera pose estimation method may be used.

The pose in the present disclosure may refer to result information obtained through visual localization, including, for example, the position and direction of a user device (eg, a device including a camera).

As another example, a pose in the present disclosure may include state information corresponding to, for example, the user's appearance and/or movements. In this example, the pose may be obtained by a deep learning module that includes an algorithm for estimating the user's pose. In this example, the pose may be obtained based on visual localization and/or inverse kinematics.

The relative pose and/or relative camera pose in the present disclosure may include pose information in a relative concept based on the pose determined by visual localization. For example, a pose determined by visual localization may be referred to as an absolute pose, and a relative coordinate value using the 3D coordinate value of the absolute pose as a reference point may correspond to a relative pose and/or a relative camera pose.

1 schematically illustrates a system for determining the validity of a camera pose using visual localization in accordance with some embodiments of the present disclosure.

Computing device 100 according to an embodiment of the present disclosure may include a processor 110 and a memory 130.

The configuration of the computing device 100 shown in FIG. 1 is only a simplified example. In one embodiment of the present disclosure, the computing device 100 may include different configurations for performing the computing environment of the computing device 100, and only some of the disclosed configurations may configure the computing device 100.

The computing device 100 in the present disclosure may refer to any type of node constituting a system for implementing embodiments of the present disclosure. Computing device 100 may refer to any type of user terminal or any type of server.

Servers may include any type of computing system or computing device, such as, for example, microprocessors, mainframe computers, digital processors, portable devices, and device controllers.

In an additional embodiment, the above-described server may include a storage unit (not shown) that stores and manages a reference image, metadata corresponding to the reference image, 3D map, etc. This storage may be included within the server or may exist under the management of the server. As another example, the storage unit may be implemented in a form that exists outside the server and can communicate with the server. In this case, the storage may be managed and controlled by an external server that is different from the server.

The components of the computing device 100 described above are exemplary and some may be excluded or additional components may be included. For example, when the above-described computing device 100 includes a user terminal, an output unit (not shown) and an input unit (not shown) may be included within the scope of the computing device 100.

Computing device 100 in this disclosure may include, for example, a digital twin server and/or a VPS server.

A user device (eg, the first device 200, etc.) in the present disclosure may refer to a device for generating query data. Information captured by the user device may include query data including a query image. The query image captured by the user device may be used for comparison with a reference image within a computing device (eg, a server), and thus the pose of the user device at the time of capture may be determined.

A user device (eg, first device 200, etc.) in this disclosure may include any type of mobile device including a camera and/or lidar. For example, the user device may include a head mounted device (HMD), augmented reality (AR) glasses, and/or virtual reality (VR) glasses.

Hereinafter, for convenience of explanation, the user device (e.g., the first device 200, etc.) and the computing device 100 are distinguished, and the computing device 100 receives shooting information from the user device, The methodology for visual localization to estimate the pose of a device will be explained. However, embodiments that perform visual localization on a user device that generates query data may also be included within the scope of the present disclosure. In this embodiment, the user device may perform at least part of the role of the computing device 100.

In one embodiment, the processor 110 may consist of at least one core, including a central processing unit (CPU) and a general purpose graphics processing unit (GPGPU) of the computing device 100. , may include a processor for data analysis and/or processing, such as a tensor processing unit (TPU).

Processor 110 may read a computer program stored in memory 130 and perform methodologies to support visual localization methodologies according to an embodiment of the present disclosure. Additionally, the processor 110 may read a computer program stored in the memory 130 and perform methods for supporting visual localization and/or tactical training according to an embodiment of the present disclosure.

According to an embodiment of the present disclosure, the processor 110 may perform an operation for learning a neural network. The processor 110 is used for learning neural networks, such as processing input data for learning in deep learning (DL), extracting features from input data, calculating errors, and updating the weights of the neural network using backpropagation. Calculations can be performed. At least one of the CPU, GPGPU, and TPU of the processor 110 may process learning of the network function. For example, CPU and GPGPU can work together to process learning of network functions and data classification using network functions. Additionally, in one embodiment of the present disclosure, processors of a plurality of computing devices may be used together to process learning of a network function and data classification using a network function. Additionally, a computer program executed in a computing device according to an embodiment of the present disclosure may be a CPU, GPGPU, or TPU executable program. Operations for a neural network according to an embodiment of the present disclosure will be described later with reference to FIG. 2.

Additionally, processor 110 may typically handle overall operations of computing device 100. For example, the processor 110 processes data, information, or signals input or output through components included in the computing device 100 or runs an application program stored in the storage to provide information or information appropriate to the user. Functions can be provided or processed.

According to one embodiment of the present disclosure, the memory 130 may store any type of information generated or determined by the processor 110 and any type of information received by the computing device 100. According to an embodiment of the present disclosure, the memory 130 may be a storage medium that stores computer software that allows the processor 110 to perform operations according to embodiments of the present disclosure. Accordingly, the memory 130 may refer to computer readable media for storing software codes required to perform embodiments of the present disclosure, data to be executed by the codes, and execution results of the codes.

According to one embodiment of the present disclosure, the memory 130 may refer to any type of storage medium. For example, the memory 130 may be a flash memory type or a hard disk type. ), multimedia card micro type, card type memory (e.g. SD or XD memory, etc.), RAM (Random Access Memory), SRAM (Static Random Access Memory), ROM (Read-Only) Memory, ROM), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM (Programmable Read-Only Memory), magnetic memory, magnetic disk, and optical disk. The computing device 100 may operate in connection with web storage that performs a storage function of the memory 130 on the Internet. The description of the memory described above is only an example, and the memory 130 used in the present disclosure is not limited to the examples described above.

The network unit 150 in the present disclosure can be configured regardless of the communication mode, such as wired or wireless, and can be used in various communication networks such as a personal area network (PAN) and a wide area network (WAN). It can be configured. In addition, the network unit 150 can operate based on the World Wide Web (WWW) and can use wireless transmission technology used for short-distance communication, such as Infrared Data Association (IrDA) or Bluetooth. It may be possible.

Computing device 100 in the present disclosure may include any type of user terminal and/or any type of server. Accordingly, embodiments of the present disclosure may be performed by a server and/or a user device.

The processor 110 of the computing device 100 may obtain the first query image corresponding to the first time from the first device 200.

The first time may be a predetermined time or a randomly selected time. For example, the first time may be 1:30, a time 100 seconds after a predetermined reference time (eg, 1:30).

The first query image may be at least one image acquired through an imaging device (eg, the first device 200) at a first time. The first query image may include at least one depth map and/or at least one of various images excluding the depth map.

The depth map is a three-dimensional value corresponding to each pixel present in an image (e.g., an image taken by the first device 200, an image taken by a virtual device (e.g., a second device, etc.), etc.) It may be an image or information expressed by distinguishing the relative distance from the image plane corresponding to the image. Accordingly, the depth map may include information related to the distance from the location where a specific image is captured to the surface of the subject (eg, an object, etc.). The depth map may include metadata corresponding to the depth map.

Metadata may refer to additional information used to perform visual localization. For example, metadata includes information related to an imaging device (e.g., a first device 200, a virtual device (e.g., a second device, etc.), etc.), pixel coordinate information of feature points, and a descriptor (descriptor) of feature points. , descriptor), location information of an object included in an image (e.g., an image taken by the first device 200, an image taken by a virtual device (e.g., a second device, etc.)), an image (e.g. For example, an image taken by the first device 200, an image taken by a virtual device (e.g., a second device, etc.), landmark information obtained from an image, text information obtained from an image, etc. Additional information may be included.

The three-dimensional value is a three-dimensional value corresponding to each pixel present in an image (e.g., an image taken by the first device 200, an image taken by a virtual device (e.g., a second device, etc.), etc.). It may be a value expressed in (three dimensions).

The image plane may correspond to the screen of a device (eg, a first device 200, a virtual device (eg, a second device, etc.)) arranged to capture an image, area, etc.

Various images other than the depth map may include, for example, 2D images, black and white images, color images, etc.

The processor 110 may obtain first comparison data by comparing the first query feature point information of the first query image obtained using a pre-learned feature point acquisition model with the reference feature point information of the pre-stored first reference image. .

Additionally, the processor 110 may match the reference image feature point information of each of the plurality of pre-stored reference images with the first query feature point information. The processor 110 may determine a matching reference image that matches the first query feature point information among a plurality of pre-stored reference images as the first reference image. For example, the processor 110 may determine the reference image of the reference image feature point information with the most feature points matching the first query feature point information as the matching reference image. And, the processor 110 may determine the matching reference image as the first reference image. The processor 110 may obtain first comparison data by comparing the first query feature point information with the reference feature point information of the first reference image.

The feature point acquisition model may include a neural network built through deep learning or machine learning. The feature point acquisition model can be trained in advance through deep learning or machine learning to acquire feature points from the query image included in the input data. The feature point acquisition model can be learned using a dataset. A dataset may refer to a set of data for performing learning and verification of a neural network. The dataset may include a training dataset and/or a validation dataset. A learning dataset may be a set of data used in the learning process of a neural network. For example, a learning dataset may be a set of data used for learning in the learning process of a feature point acquisition model. A validation dataset may be a set of data used to evaluate a neural network. For example, a validation dataset may be a set of data used to evaluate a model for visual localization. As another example, the feature point acquisition model may correspond to a pre-trained model based on supervised learning to recognize as a feature point a point where the amount of change in at least one of color or geometric patterns in the object exceeds a predetermined threshold.

The first query feature point information may be information related to a feature acquired from the first query image input to the pre-learned feature point acquisition model. For example, the first query feature point information may include at least one first query feature point and/or a first query descriptor (descriptor). At least one first query feature point may be a coordinate for each feature part of the first query image. At least one first query descriptor may include information for describing each of at least one first query feature point. For example, at least one first query descriptor may include information about the directionality and size of each of the at least one first query feature point and/or the relationship between pixels surrounding each of the at least one first query feature point.

The first reference image may be a pre-stored image used to estimate the pose of the first device 200 that captured the first query image.

The reference feature point information of the first reference image may be information related to a feature obtained from the first reference image. For example, reference feature point information of the first reference image may include at least one first reference feature point and/or a first reference descriptor. At least one first reference feature point may be a coordinate for each feature part in the first reference image. At least one first reference descriptor may include information for describing each of at least one first reference feature point. For example, the at least one first reference descriptor may include information about the directionality and size of each of the at least one first reference feature point and/or the relationship between pixels surrounding each of the at least one first reference feature point.

The first comparison data may include information about each of at least one matching feature point (or a plurality of matching feature points) that matches each other by comparing the first query feature point information with the reference feature point information of the first reference image. For example, the first comparison data may include 2D matching coordinate values and/or 3D matching coordinate values of each of at least one matching feature point. The 2D matching coordinate value may include the two-dimensional coordinates of the matching feature point (e.g., (x,y), where x and y are real numbers, etc.). For example, the 2D matching coordinate value may include two-dimensional pixel coordinate values in the first query image. The 3D matching coordinate value may include the three-dimensional coordinates of the matching feature point (e.g., (x, y, z), where x, y, and z are real numbers, etc.). For example, the 3D matching coordinate value may include three-dimensional coordinate values in the coordinate system of the first reference device associated with the pre-stored first reference image.

The processor 110 may estimate the first camera pose of the first device 200 based on the first comparison data. For example, the processor 110 may estimate the first camera pose based on the 2D matching coordinate values and 3D matching coordinate values of each of the plurality of matching feature points included in the first comparison data.

The first camera pose may include location information and direction information of the first device 200 related to the first query image. The location information of the first camera pose may include coordinates (e.g., (x,y,z), where x, y, and z are real numbers, etc.) where the first device 200 exists. . The direction information of the first camera pose may include at least one of the direction and/or angle from which the first device 200 is viewed.

For example, the processor 110 may obtain a second query image corresponding to a second time different from the first time from the first device 200.

The second time may be a predetermined time different from the first time or a randomly selected time. For example, the second time may be 2:30, a time 100 seconds after a predetermined reference time (eg, 2:30).

For another example, the processor 110 may receive a second query image from a second device arranged based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device 200. It can be obtained.

The inertial measurement unit is included in the first device 200 and includes an acceleration sensor that measures the direction of movement, distance and speed associated with the linear motion of the first device 200 and an angular velocity associated with the rotational motion of the first device 200. It may include a gyro sensor that measures. The inertial measurement unit measures the direction of movement, distance and speed associated with the linear motion of each of the X, Y and Z axes, and the rotational motion of each of the The movement direction, movement distance, inclination, etc. of the first device 200 may be obtained based on the acceleration and gyro sensing data obtained by further measuring the angular velocity. However, the inertial measurement unit is not limited to this and may include various measurement units capable of obtaining information related to movement.

The information of the inertial measurement unit may include information related to the movement of the first device 200 from the first time to the second time. For example, the information of the inertial measurement unit is at least one of the movement direction, movement distance, and/or inclination of the first device 200 obtained through the inertial measurement unit of the first device 200 from the first time to the second time. It can contain one.

The second query image is at least one image acquired through an imaging device (e.g., a first device 200, a virtual device (e.g., a second device, etc.), etc.) at a second time different from the first time. It can be. The second query image may include at least one depth map and/or at least one of various images excluding the depth map.

Processor 110 may create a 3D model. For example, the processor 110 may acquire spatial information about an arbitrary space scanned or photographed using an imaging device (eg, a space related to an image received from the first device 200).

An imaging device may refer to any type of equipment for detecting optical images, converting them into electrical signals, and inputting them into the computing device 100. For example, the imaging device may include at least one of a scanner, Lidar, and/or a vision sensor. The computing device 100 may include an imaging device or may be linked to an external imaging device wirelessly or wired.

Spatial information may refer to information about the interior of a specific space. For example, spatial information may mean information about any type of object existing in any space. For example, spatial information may include at least one of distance, direction, and/or speed from an imaging device to an object existing in a certain space. Additionally, the spatial information may include at least one of the characteristics of the object's color, temperature, material distribution, and/or concentration. An object may refer to a software module that includes state data representing a state and operation (eg, procedure, method, function, etc.) data related to the state data. For example, the object may include at least one of a basic object, a moving object, and/or a specific predetermined object.

The processor 110 may generate point cloud information including color data and/or location data based on spatial information.

Point cloud information may refer to information representing a set of points in a three-dimensional space (eg, 3D model, etc.). Point cloud information may refer to location information of measurement points related to the color and/or location of a basic object existing in an arbitrary space. The basic object may be an object excluding a moving object and a predetermined specific object. Accordingly, the point cloud information may be information from which information related to at least one of lighting, a moving object, and/or a predetermined specific object has been removed. Lighting information may be information related to light. A mobile object may be an object that is not permanently fixed but can move at a specific point in time or in a specific situation. For example, a moving object may be an object that can move (eg, a car, a bicycle, etc.). The specific predetermined object may be a floating object (eg, a plant) rather than a moving object. According to one embodiment, the predetermined specific object may be an object set by the user.

Additionally, point cloud information may include color data and/or location data. Color data may include data related to the color of an object existing in an arbitrary space. For example, color data may include vertex color values that include RGB color values of objects existing in arbitrary space. However, color data is not limited to this and may include various values representing colors.

Location data may include data related to the location of an object existing in arbitrary space. For example, location data may include the x, y, and z location coordinates of an object existing in arbitrary space. However, the meaning of location data is not limited to the above-described example, and may include various values indicating the location or depth of an object.

The processor 110 may generate a 3D model based on point cloud information.

A 3D model is a model created in three dimensions to correspond to an arbitrary space based on point cloud information, and may include point cloud information. For example, the processor 110 may change the point cloud included in the point cloud information in mesh units to generate a 3D model formed in three dimensions corresponding to an arbitrary space. Accordingly, the processor 110 may generate a 3D model by changing the point cloud included in the point cloud information into various other units. For another example, the processor 110 may generate a 3D model formed from a point cloud included in point cloud information. Accordingly, the processor 110 can generate a 3D model by using the point cloud included in the point cloud information without changing it. Here, the arbitrary space may be a space scanned or photographed using an imaging device, and may be a space related to the image received from the first device 200.

The second device may include a virtual device placed on the 3D model. The virtual device may include a camera for photographing a specific area. The processor 110 may determine location information and direction information of the second device based on the first camera pose and information on the inertial measurement unit included in the first device 200. The location information of the second device may include coordinates at which the second device exists (eg, coordinates on a 3D model of the second device). The direction information of the second device may include at least one of a direction and/or angle from which the second device is looking (eg, a direction and/or angle on a 3D model of the second device).

The processor 110 may place the second device in the 3D model based on the location information and direction information of the second device. When the processor 110 obtains camera information of the first device 200 from the first device 200, the processor 110 sets the second device settings applied to the 3D model environment based on the camera information of the first device 200. (For example, camera specifications, etc.) can be determined.

The camera information of the first device 200 may be information about a camera for photographing a specific area. For example, camera information of the first device 200 may include at least one of camera location information and/or direction information. Additionally, the camera information of the first device 200 may include hardware-related information of the first device 200. For example, the camera information of the first device 200 includes at least one of focal length information with a specific area, principal point information, lens information, and/or 'sensor and/or film information'. It can be included. The focal length information may be information about the distance from the main point of the first device 200 to a specific area to be photographed. Principal point information may be information about a point where a principal plane intersects an optical axis. Lens information may be information related to distortion of a lens applied to the camera of the first device 200. The film information may be information related to the size of the film applied to the camera of the first device 200. The sensor information may be information related to a camera sensor applied to the camera of the first device 200. The processor 110 may determine the aspect ratio (ratio of width to height) of an image acquired through a virtual camera based on sensor or film information of the first device 200. However, the camera information of the first device 200 is not limited to this and may include various camera information.

The processor 110 may obtain second comparison data by comparing second query feature point information and first query feature point information of the second query image obtained using a pre-learned feature point acquisition model.

The second query feature point information may be information related to features obtained from the second query image input to the pre-learned feature point acquisition model. For example, the second query feature point information may include at least one second query feature point and/or a second query descriptor. At least one second query feature point may be a coordinate for each feature part of the second query image. At least one second query descriptor may include information for describing each of at least one second query feature point. For example, at least one second query descriptor may include information about the directionality and size of each of the at least one second query feature point and/or the relationship between pixels surrounding each of the at least one second query feature point.

The second comparison data may include information about each of at least one matching feature point (or a plurality of matching feature points) that matches each other by comparing the first query feature point information and the second query feature point information. For example, the second comparison data may include 2D matching coordinate values and/or 3D matching coordinate values of each of at least one matching feature point. The 2D matching coordinate value may include the two-dimensional coordinates of the matching feature point (e.g., (x,y), where x and y are real numbers, etc.). The 3D matching coordinate value may include the three-dimensional coordinates of the matching feature point (e.g., (x, y, z), where x, y, and z are real numbers, etc.). According to some embodiments of the present disclosure, the second comparison data compares reference feature point information (e.g., first reference feature point information, etc.) and second query feature point information to generate at least one matching feature point (or, (a plurality of matching feature points) may include information about each. Therefore, according to some embodiments of the present disclosure, the processor 110 may obtain second comparison data by comparing the second query feature point information and reference feature point information (e.g., first reference feature point information, etc.) .

The processor 110 may detect the first device 200 related to the second query image, for example, based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device 200. ) can estimate the second camera pose.

For another example, the processor 110 may use the first camera pose and the second camera pose of the second device arranged based on the information of the inertial measurement unit (IMU) included in the first device 200. It can be estimated.

The second camera pose may include location information and direction information of the first device 200 or the second device related to the second query image. The location information of the second camera pose includes coordinates (e.g., (x,y,z), where x, y and z are real numbers, etc.) where the first device 200 or the second device exists. It can be included. The direction information of the first camera pose may include at least one of the direction and/or angle from which the first device 200 or the second device is looking.

The processor 110 may determine an effective value of at least one of the first camera pose and the second camera pose based on the first comparison data, the second comparison data, and the information of the inertial measurement unit.

For example, the processor 110 may estimate a relative camera pose between the first device and the first reference device associated with the pre-stored first reference image, based on the first comparison data.

The processor 110 may obtain the first query 3D estimated coordinate value of the first query image based on the relative camera pose and the first reference 3D coordinate value of the pre-stored first reference image.

The processor 110 may obtain the second query 3D estimated coordinate value of the second query image based on the first query 3D estimated coordinate value and information of the inertial measurement unit.

The processor 110 may obtain the second query 2D estimated coordinate value of the second query image from the second query 3D estimated coordinate value using a predetermined algorithm.

The processor 110 may obtain the second query 2D actual coordinate value of the second query image based on the second comparison data.

The processor 110 may determine the value of the reprojection error calculated by comparing the second query 2D estimated coordinate value and the second query 2D actual coordinate value as a valid value of at least one of the first camera pose and the second camera pose.

A method of determining a valid value of at least one of the first camera pose or the second camera pose according to some embodiments of the present disclosure will be described later with reference to FIGS. 3 and 4.

In an additional embodiment, the computing device 100 may determine a pose estimation algorithm used to perform visual localization among a plurality of pose estimation algorithms based on camera information of the first device 200. For example, pose estimation for the first device 200 may estimate 6DoF (Six degrees of freedom) in a direction that minimizes reprojection error. 6DoF can be the six rotational and translational motion components of an object moving in three-dimensional space.

For example, when the camera's internal parameters are not given (i.e., an uncalibrated camera), the pose estimation algorithm uses the Direct Linear Transformation (DLT) methodology to calculate the elements of the rotation matrix, the translation vector ( The reprojection error equation, which includes the elements of translation vector and the camera's internal parameters as unknowns, can be changed to a linear equation and the rotation matrix, translation vector, and camera matrix can be estimated through QR decomposition.

For example, when the camera's internal parameters are given (i.e., a calibrated camera), the pose estimation algorithm uses the Perspective-n-Point (PnP) methodology to transform the reprojection error, a non-linear equation, into a linear equation. Differently, pose estimation can be performed by directly solving the reprojection error, which is a non-linear equation, using the Gauss-Newton methodology or Levenberg-Marquardt methodology. In this example, representative PnP methodologies may include P3P, EPnP, and SQPnP methodologies.

In a further example, the DLT and PnP methodologies described above may introduce RANSAC to minimize camera pose errors resulting from inaccuracies in the 3D coordinates of reference feature points or inaccuracies in feature points.

When the effective value of at least one of the first camera pose and the second camera pose corresponds to a predetermined condition, the processor 110 may re-estimate at least one of the first camera pose and the second camera pose. A valid value of at least one of the first camera pose and the second camera pose may be a value of reprojection error. However, the at least one valid value is not limited to this. The predetermined condition may be a case where the effective value of at least one of the first camera pose or the second camera pose is greater than (or less than, less than, greater than, etc.) a certain threshold value.

For example, the processor 110 may re-acquire the first query image corresponding to the first time from the first device 200 described above. The processor 110 may re-acquire first comparison data by comparing the first query feature point information of the first query image obtained using a pre-learned feature point acquisition model with the reference feature point information of the pre-stored first reference image. there is. The processor 110 may re-estimate the first camera pose of the first device 200 based on the first comparison data. The processor 110 may re-acquire a second query image corresponding to a second time different from the first time from the first device 200. The processor 110 may re-acquire second comparison data by comparing second query feature point information and first query feature point information of the second query image obtained using a pre-learned feature point acquisition model. The processor 110 creates a second camera pose of the first device 200 related to the second query image based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device. It can be re-estimated. The processor 110 may re-determine the effective value of at least one of the first camera pose and the second camera pose based on the first comparison data, the second comparison data, and the information of the inertial measurement unit. However, the re-estimation process performed through the processor 110 described above is only a simplified example. In one embodiment of the present disclosure, the processor 110 may include other processes for performing re-estimation, and only some of the above-described re-estimation processes may be performed.

The first device 200 may include any type of terminal capable of interacting with a server or other computing device. The first device 200 includes, for example, mobile phones, smart phones, laptop computers, personal digital assistants (PDAs), slate PCs, tablet PCs, and ultrabooks. (ultrabook) may be included.

Additionally, the first device 200 may include software and/or hardware for generating images. Accordingly, the first device 200 may generate an image and transmit it to the computing device 100. In one embodiment, the first device 200 may refer to any type of equipment for detecting an optical image, converting it into an electrical signal, and transmitting the signal to the computing device 100. For example, the first device 200 may include at least one of a camera, scanner, Lidar, and/or vision sensor. The computing device 100 may include a device (eg, the first device 200) or may be linked to an external device (eg, the first device 200) wirelessly or wired.

The network may include any wired or wireless communication network through which the computing device 100 and the first device 200 can transmit and receive data and signals of any type to each other. For example, the computing device 100 may receive a query image from the first device 200 through a network.

Figure 2 is a schematic diagram showing a network function according to an embodiment of the present disclosure.

In this disclosure, the feature point acquisition model may correspond to an artificial intelligence-based model. In additional embodiments of the present disclosure, artificial intelligence-based models may be used to estimate the pose of a camera (e.g., first camera pose, second camera pose, etc.) and/or create a character.

Throughout this disclosure, the terms artificial intelligence-based model, model, computational model, neural network, network function, and neural network may be used interchangeably. A neural network can generally consist of a set of interconnected computational units, which can be referred to as nodes. These nodes may also be referred to as neurons. A neural network consists of at least one node. Nodes (or neurons) that make up neural networks may be interconnected by one or more links.

Within a neural network, one or more nodes connected through a link may form a relative input node and output node relationship. The concepts of input node and output node are relative, and any node in an output node relationship with one node may be in an input node relationship with another node, and vice versa. As described above, input node to output node relationships can be created around links. One or more output nodes can be connected to one input node through a link, and vice versa.

In a relationship between an input node and an output node connected through one link, the value of the data of the output node may be determined based on the data input to the input node. Here, the link connecting the input node and the output node may have a weight. Weights may be variable and may be varied by the user or algorithm in order for the neural network to perform the desired function. For example, when one or more input nodes are connected to one output node by respective links, the output node is set to the values input to the input nodes connected to the output node and the links corresponding to each input node. The output node value can be determined based on the weight.

As described above, in a neural network, one or more nodes are interconnected through one or more links to form an input node and output node relationship within the neural network. The characteristics of the neural network can be determined according to the number of nodes and links within the neural network, the correlation between the nodes and links, and the value of the weight assigned to each link. For example, if the same number of nodes and links exist and two neural networks with different weight values of the links exist, the two neural networks may be recognized as different from each other.

A neural network may consist of a set of one or more nodes. A subset of nodes that make up a neural network can form a layer. Some of the nodes constituting the neural network may form one layer based on the distances from the first input node. For example, a set of nodes with a distance n from the initial input node may constitute n layers. The distance from the initial input node can be defined by the minimum number of links that must be passed to reach the node from the initial input node. However, this definition of a layer is arbitrary for explanation purposes, and the order of a layer within a neural network may be defined in a different way than described above. For example, a layer of nodes may be defined by distance from the final output node.

The initial input node may refer to one or more nodes in the neural network through which data is directly input without going through links in relationships with other nodes. Alternatively, in a neural network network, in the relationship between nodes based on links, it may mean nodes that do not have other input nodes connected by links. Similarly, the final output node may refer to one or more nodes that do not have an output node in their relationship with other nodes among the nodes in the neural network. Additionally, hidden nodes may refer to nodes constituting a neural network other than the first input node and the last output node.

The neural network according to an embodiment of the present disclosure is a neural network in which the number of nodes in the input layer may be the same as the number of nodes in the output layer, and the number of nodes decreases and then increases again as it progresses from the input layer to the hidden layer. You can. In addition, the neural network according to another embodiment of the present disclosure may be a neural network in which the number of nodes in the input layer may be less than the number of nodes in the output layer, and the number of nodes decreases as it progresses from the input layer to the hidden layer. there is. In addition, the neural network according to another embodiment of the present disclosure may be a neural network in which the number of nodes in the input layer may be greater than the number of nodes in the output layer, and the number of nodes increases as it progresses from the input layer to the hidden layer. You can. A neural network according to another embodiment of the present disclosure may be a neural network that is a combination of the above-described neural networks.

A deep neural network (DNN) may refer to a neural network that includes multiple hidden layers in addition to the input layer and output layer. Deep neural networks allow you to identify latent structures in data. In other words, it is possible to identify the potential structure of a photo, text, video, voice, or music (e.g., what object is in the photo, what the content and emotion of the text are, what the content and emotion of the voice are, etc.) . Deep neural networks include convolutional neural network (CNN), recurrent neural network (RNN), auto encoder, restricted Boltzmann machine (RBM), and deep trust network ( It may include deep belief network (DBN), Q network, U network, Siamese network, generative adversarial network (GAN), etc. The description of the deep neural network described above is only an example and the present disclosure is not limited thereto.

A neural network may be trained in at least one of supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning. Learning of a neural network may be a process of applying knowledge for the neural network to perform a specific operation to the neural network.

Neural networks can be trained to minimize output errors. In neural network learning, learning data is repeatedly input into the neural network, the output of the neural network and the error of the target for the learning data are calculated, and the error of the neural network is transferred from the output layer of the neural network to the input layer in the direction of reducing the error. This is the process of updating the weight of each node in the neural network through backpropagation. In the case of supervised learning, learning data in which the correct answer is labeled for each learning data is used (i.e., labeled learning data), while in the case of unsupervised learning, the correct answer may not be labeled in each learning data. That is, for example, in the case of teacher learning regarding data classification, the learning data may be data in which each learning data is labeled with a category. Labeled training data is input to the neural network, and the error can be calculated by comparing the output (category) of the neural network with the label of the training data. As another example, in the case of unsupervised learning on data classification, the error can be calculated by comparing the input training data with the neural network output. The calculated error is backpropagated in the reverse direction (i.e., from the output layer to the input layer) in the neural network, and the connection weight of each node in each layer of the neural network can be updated according to backpropagation. The amount of change in the connection weight of each updated node may be determined according to the learning rate. The neural network's calculation of input data and backpropagation of errors can constitute a learning cycle (epoch). The learning rate may be applied differently depending on the number of repetitions of the learning cycle of the neural network. For example, in the early stages of neural network training, a high learning rate can be used to increase efficiency by allowing the neural network to quickly achieve a certain level of performance, and in the later stages of training, a low learning rate can be used to increase accuracy.

In the learning of neural networks, the training data can generally be a subset of real data (i.e., the data to be processed using the learned neural network), and thus the error for the training data is reduced, but the error for the real data is reduced. There may be an incremental learning cycle. Overfitting is a phenomenon in which errors in actual data increase due to excessive learning on training data. For example, a phenomenon in which a neural network that learned a cat by showing a yellow cat fails to recognize that it is a cat when it sees a non-yellow cat may be a type of overfitting. Overfitting can cause errors in machine learning algorithms to increase. To prevent such overfitting, various optimization methods can be used. To prevent overfitting, methods such as increasing the learning data, regularization, dropout to disable some of the network nodes during the learning process, and use of a batch normalization layer can be applied. You can.

A computer-readable medium storing a data structure according to an embodiment of the present disclosure is disclosed.

Data structure can refer to the organization, management, and storage of data to enable efficient access and modification of data. Data structure can refer to the organization of data to solve a specific problem (e.g., retrieving data, storing data, or modifying data in the shortest possible time). A data structure may be defined as a physical or logical relationship between data elements designed to support a specific data processing function. Logical relationships between data elements may include connection relationships between user-defined data elements. Physical relationships between data elements may include actual relationships between data elements that are physically stored in a computer-readable storage medium (e.g., a persistent storage device). A data structure may specifically include a set of data, relationships between data, and functions or instructions applicable to the data. Effectively designed data structures allow computing devices to perform computations while minimizing the use of the computing device's resources. Specifically, computing devices can increase the efficiency of operations, reading, insertion, deletion, comparison, exchange, and search through effectively designed data structures.

Data structures can be divided into linear data structures and non-linear data structures depending on the type of data structure. A linear data structure may be a structure in which only one piece of data is connected to another piece of data. Linear data structures may include List, Stack, Queue, and Deque. A list can refer to a set of data that has an internal order. The list may include a linked list. A linked list may be a data structure in which data is connected in such a way that each data is connected in a single line with a pointer. In a linked list, a pointer may contain connection information to the next or previous data. Depending on its form, a linked list can be expressed as a singly linked list, a doubly linked list, or a circularly linked list. A stack may be a data listing structure that allows limited access to data. A stack can be a linear data structure in which data can be processed (for example, inserted or deleted) at only one end of the data structure. Data stored in the stack may have a data structure (LIFO-Last in First Out) where the later it enters, the sooner it comes out. A queue is a data listing structure that allows limited access to data. Unlike the stack, it can be a data structure (FIFO-First in First Out) where data stored later is released later. A deck can be a data structure that can process data at both ends of the data structure.

A non-linear data structure may be a structure in which multiple pieces of data are connected behind one piece of data. Nonlinear data structures may include graph data structures. A graph data structure can be defined by vertices and edges, and an edge can include a line connecting two different vertices. Graph data structure may include a tree data structure. A tree data structure may be a data structure in which there is only one path connecting two different vertices among a plurality of vertices included in the tree. In other words, it may be a data structure that does not form a loop in the graph data structure.

Throughout this specification, computational model, neural network, network function, and neural network may be used interchangeably. Below, it is described in a unified manner as a neural network. Data structures may include neural networks. And the data structure including the neural network may be stored in a computer-readable medium. Data structures including neural networks also include data preprocessed for processing by a neural network, data input to the neural network, weights of the neural network, hyperparameters of the neural network, data acquired from the neural network, activation functions associated with each node or layer of the neural network, neural network It may include a loss function for learning. A data structure containing a neural network may include any of the components disclosed above. In other words, the data structure including the neural network includes preprocessed data for processing by the neural network, data input to the neural network, weights of the neural network, hyperparameters of the neural network, data acquired from the neural network, activation functions associated with each node or layer of the neural network, neural network It may be configured to include all or any combination of the loss function for learning. In addition to the configurations described above, a data structure containing a neural network may include any other information that determines the characteristics of the neural network. Additionally, the data structure may include all types of data used or generated in the computational process of a neural network and is not limited to the above. Computer-readable media may include computer-readable recording media and/or computer-readable transmission media. A neural network can generally consist of a set of interconnected computational units, which can be referred to as nodes. These nodes may also be referred to as neurons. A neural network consists of at least one node.

The data structure may include data input to the neural network. A data structure containing data input to a neural network may be stored in a computer-readable medium. Data input to the neural network may include learning data input during the neural network learning process and/or input data input to the neural network on which training has been completed. Data input to the neural network may include data that has undergone pre-processing and/or data subject to pre-processing. Preprocessing may include a data processing process to input data into a neural network. Therefore, the data structure may include data subject to preprocessing and data generated by preprocessing. The above-described data structure is only an example and the present disclosure is not limited thereto.

The data structure may include the weights of the neural network. (In this specification, weights and parameters may be used with the same meaning.) And the data structure including the weights of the neural network may be stored in a computer-readable medium. A neural network may include multiple weights. Weights may be variable and may be varied by the user or algorithm in order for the neural network to perform the desired function. For example, when one or more input nodes are connected to one output node by respective links, the output node is set to the values input to the input nodes connected to the output node and the links corresponding to each input node. Based on the weight, the data value output from the output node can be determined. The above-described data structure is only an example and the present disclosure is not limited thereto.

As an example and not a limitation, the weights may include weights that are changed during the neural network learning process and/or weights for which neural network learning has been completed. Weights that change during the neural network learning process may include weights that change at the start of the learning cycle and/or weights that change during the learning cycle. Weights for which neural network training has been completed may include weights for which a learning cycle has been completed. Therefore, the data structure including the weights of the neural network may include weights that are changed during the neural network learning process and/or the data structure including the weights for which neural network learning has been completed. Therefore, the above-mentioned weights and/or combinations of each weight are included in the data structure including the weights of the neural network. The above-described data structure is only an example and the present disclosure is not limited thereto.

The data structure including the weights of the neural network may be stored in a computer-readable storage medium (e.g., memory, hard disk) after going through a serialization process. Serialization can be the process of converting a data structure into a form that can be stored on the same or a different computing device and later reorganized and used. Computing devices can transmit and receive data over a network by serializing data structures. Data structures containing the weights of a serialized neural network can be reconstructed on the same computing device or on a different computing device through deserialization. The data structure including the weights of the neural network is not limited to serialization. Furthermore, the data structure including the weights of the neural network is a data structure to increase computational efficiency while minimizing the use of computing device resources (e.g., in non-linear data structures, B-Tree, Trie, m-way search tree, AVL tree, Red-Black Tree) may be included. The foregoing is merely an example and the present disclosure is not limited thereto.

The data structure may include hyper-parameters of a neural network. And the data structure including the hyperparameters of the neural network can be stored in a computer-readable medium. A hyperparameter may be a variable that can be changed by the user. Hyperparameters include, for example, learning rate, cost function, number of learning cycle repetitions, weight initialization (e.g., setting the range of weight values subject to weight initialization), Hidden Unit. It may include a number (e.g., number of hidden layers, number of nodes in hidden layers). The above-described data structure is only an example and the present disclosure is not limited thereto.

FIG. 3 is a diagram illustrating a process for estimating a camera pose by a computing device according to some embodiments of the present disclosure.

Referring to FIG. 3, the processor 110 of the computing device 100 selects a first reference image associated with the first device 331 and the first reference image 322 corresponding to the first time, based on the first comparison data. Relative camera poses (R _re , T _re ) may be estimated and/or obtained between the devices 321 .

The processor 110 may obtain the first query 3D estimated coordinate value of the first query image based on the relative camera pose and the first reference 3D coordinate value of the pre-stored first reference image.

For example, the processor ₁₁₀ may generate relative camera poses (R _{re ,} T _re ) and ( _r , Z _r ), the first query 3D estimated coordinate value of the first query image 332 in the camera coordinate system defined based on the first device 331 corresponding to the first time can be obtained. .

_The processor ₁₁₀ calculates the second _query based on the first _query 3D _estimated coordinate values ( _coordinates ( The second query 3D estimated coordinate value of the query image can be obtained.

The processor 110 may obtain the second query 2D estimated coordinate value of the second query image 342 from the second query 3D estimated coordinate value using a predetermined algorithm. The predetermined algorithm may include an algorithm for converting three-dimensional coordinates into two-dimensional coordinates. The predetermined algorithm may include Equation 1:

x ₂ and y ₂ may be second query 2D estimated coordinate values. f _x and f _y may be a factor related to the focal length with a specific area. r may be a skew parameter. The skew parameter may be a factor related to distortion. c _x and c _y may be a factor regarding the principal point with a specific area. X ₂ , Y ₂ and Z ₂ is The second query may be a 3D estimated coordinate value. may be a matrix of a camera included in the first device. However, the predetermined algorithm is not limited to this and may include an algorithm that converts various three-dimensional coordinates into two-dimensional coordinates.

The processor 110 may obtain the second query 2D actual coordinate value of the second query image based on the second comparison data.

The processor 110 may determine the value of the reprojection error calculated by comparing the second query 2D estimated coordinate value and the second query 2D actual coordinate value as a valid value of at least one of the first camera pose and/or the second camera pose. there is.

Reprojection error may mean the difference or difference value between the estimated coordinate value and the actual coordinate value. For example, reprojection error may be the distance between the pixel value estimated through calculation and the actual pixel value.

According to some embodiments of the present disclosure, the processor 110 may obtain the relative camera pose between the first reference device 321 and the first device 331 corresponding to the first time from Equation 2 below.

With respect to the point P (310) in a three-dimensional space (e.g., real space, virtual space, 3D model space, etc.) that matches each other based on the first comparison data, based on the first reference device 321 The coordinates of point P _r (323) in the defined camera coordinate system are (X _r , Y _r , Z _r ), and the point P in the camera coordinate system defined based on the first device 331 corresponding to the first time The coordinates of ₁ (333) may be (X ₁ , Y ₁ , Z ₁ ). The relative camera pose between the first device 331 corresponding to the first time and the first reference device 321 related to the first reference image 322 may be the relative camera pose (R _re , T _re ). R _re is the first reference _, so _that it corresponds _{to (} It may be a value multiplied by (X _r , Y _r , Z _r ), which are the coordinates of the point P _r (323) in the camera coordinate system defined based on the device 321. T _re is the first reference _, so _that it corresponds _{to (} It may be a value that is additionally added to (X _r , Y _r , Z _r ), which are the coordinates of the point P _r (323) in the camera coordinate system defined based on the device 321.

According to some embodiments of the present disclosure, processor 110 may obtain a relative camera pose.

For example, the processor 110 calculates 3 in the camera coordinate system of the first reference device 321 for each of n points (where n is a natural number) based on the first device 200 and the reference feature point information. Dimensional coordinates (e.g., (x, y, z), where x, y and z are real numbers, etc.) and two-dimensional coordinates in the image captured by the first device 331 (e.g. For example, (x,y), where x and y are real numbers, etc.) can be obtained.

The processor 110 projects each of the three-dimensional coordinates of the first reference device 321 into the camera image of the first device 331 corresponding to the first time based on a plurality of candidate relative camera poses. Thus, pixel coordinate estimation values can be obtained.

The processor 110 may calculate the sum of the reprojection errors obtained by comparing the estimated pixel coordinate values and the actual pixel coordinate values, respectively.

The processor 110 selects the first candidate relative camera pose with the minimum sum of reprojection errors among the plurality of candidate relative camera poses as the first candidate relative camera pose associated with the first device 331 and the first reference image 322 corresponding to the first time. 1 The relative camera poses (R _re , T _re ) between the reference devices 321 can be determined.

FIG. 4 is a diagram illustrating a method for minimizing reprojection error of a camera pose by a computing device according to some embodiments of the present disclosure.

If the two variables x and y are in a linear relationship, the processor 110 generates data samples (e.g., {(x ₁ , y ₁ ), ~, (x _n , y _n )}, where n is a natural number, etc. It can be obtained. x of the data sample may be the two-dimensional coordinate of the query image information. y of the data sample may be the three-dimensional coordinates of reference feature point information. Data samples can be obtained from external devices. Data samples may be stored in advance in the memory 130.

The processor 110 may classify data samples into inliers and outliers.

For example, the processor 110 may select N (<n) samples in a predetermined manner (eg, randomly extracted, extracted by a predetermined number, etc.).

The processor 110 may calculate the values of a and b as A and B in the estimation equation y=ax+b of the N samples selected based on an estimation algorithm. The estimation algorithm may include the least square method. The least squares method is to minimize the sum of the squares of d _i when the distance between the estimated relationship y=ax+b and (x _i , y _i ) (where i is a natural number between 1 and n) is d _i . This may be a method of calculating a and b. According to some embodiments of the present disclosure, the values of a and b of the estimation relation y=ax+b may correspond to R _re and T _re of the relative camera pose, respectively. Estimation algorithms may include Direct Linear Transform, Levenberg-Marquardt algorithm, Gauss-Newton algorithm, PnP method, etc.

The processor 110 may calculate the estimated relationship y=Ax+B and the distance between data samples, respectively. According to some embodiments of the present disclosure, the distance between the estimation relation y=Ax+B and data samples may correspond to the reprojection error.

The processor 110 may classify data samples whose distance between the calculated estimation relation y=Ax+B and the data samples is greater than (or greater than) a predetermined specific value (e.g., 1, 2, etc.) as outliers. .

The processor 110 may classify data samples whose distance between the calculated estimation relation y=Ax+B and the data samples is less than (or less than) a predetermined specific value (e.g., 1, 2, etc.) as inliers. .

The processor 110 repeats the process of classifying the above-described data samples into inliers and outliers a predetermined number of times (e.g., 1 time, 2 times, m times (where m is a natural number), etc.) to determine the number of inliers. In most cases, A and B can be determined as the final parameter values.

For example, FIGS. 4(a) and 4(b) may have identical samples. The estimated relational equation 410 in FIG. 4(a) may have 4 samples corresponding to the inlier 420 and 5 samples corresponding to the outlier 430. The estimated relational equation 440 in FIG. 4(b) may have 7 samples corresponding to the inlier 450 and 2 samples corresponding to the outlier 430. In this case, the processor 110 may determine A and B of the estimation equation 440 of FIG. 4(b), which has a larger number of inliers, as the final parameter values among the two estimation equations 410 and 440. For example, the processor 110 determines A in the estimation equation 440 in FIG. 4(b) as the value of R _re of the relative camera pose, and sets B in the estimation equation 440 in FIG. 4(b) to T It can be determined by the value of _re .

5 and 6 exemplarily illustrate a method for determining an effective value of a camera pose according to some embodiments of the present disclosure.

Referring to FIG. 5, the processor 110 may obtain the first query image corresponding to the first time from the first device 200 (S110).

The processor 110 may obtain first comparison data by comparing the first query feature point information of the first query image obtained using a pre-learned feature point acquisition model with the reference feature point information of the pre-stored first reference image. (S120).

The pre-trained feature point acquisition model can be trained in advance through deep learning or machine learning to acquire feature points from a specific image included in specific input data.

The processor 110 may match reference image feature point information of each of the plurality of pre-stored reference images with first query feature point information. The processor 110 may determine a matching reference image that matches the first query feature point information among a plurality of pre-stored reference images as the first reference image. The processor 110 may obtain first comparison data by comparing the first query feature point information with the reference feature point information of the first reference image.

The processor 110 may estimate the first camera pose of the first device 200 based on the first comparison data (S130).

The processor 110 may estimate the first camera pose based on the 2D matching coordinate values and 3D matching coordinate values of each of the plurality of matching feature points included in the first comparison data.

The 2D matching coordinate value may include two-dimensional pixel coordinate values in the first query image.

The 3D matching coordinate value may include three-dimensional coordinate values in the coordinate system of the first reference device related to the pre-stored first reference image.

The processor 110 may obtain a second query image corresponding to a second time different from the first time from the first device 200 (S140).

The processor 110 may obtain the second query image from a second device arranged based on the first camera pose and information of the inertial measurement unit.

The processor 110 may obtain second comparison data by comparing the second query feature point information and the first query feature point information of the second query image obtained using a pre-learned feature point acquisition model (S150).

The processor 110 creates a second camera pose of the first device related to the second query image based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device 200. It can be estimated (S160).

The information of the inertial measurement unit may include information related to the movement of the first device 200 from the first time to the second time.

The processor 110 may determine the effective value of at least one of the first camera pose and the second camera pose based on the first comparison data, the second comparison data, and the information of the inertial measurement unit (S170).

When the effective value of at least one of the first camera pose and the second camera pose corresponds to a predetermined condition, the processor 110 may re-estimate at least one of the first camera pose and the second camera pose.

Referring to FIG. 6, the processor 110 may estimate a relative camera pose between the first device and a first reference device related to the pre-stored first reference image based on the first comparison data (S171). .

The processor 110 may obtain the first query 3D estimated coordinate value of the first query image based on the relative camera pose and the first reference 3D coordinate value of the pre-stored first reference image (S172).

The processor 110 may obtain the second query 3D estimated coordinate value of the second query image based on the first query 3D estimated coordinate value and the information of the inertial measurement unit (S173).

The processor 110 may obtain the second query 2D estimated coordinate value of the second query image from the second query 3D estimated coordinate value using a predetermined algorithm (S174).

The processor 110 may obtain the second query 2D actual coordinate value of the second query image based on the second comparison data (S175).

The processor 110 determines the value of the reprojection error calculated by comparing the second query 2D estimated coordinate value and the second query 2D actual coordinate value as a valid value of at least one of the first camera pose and the second camera pose. (S176).

The steps shown in FIGS. 5 and 6 are exemplary steps. Accordingly, it will also be apparent to those skilled in the art that some of the steps in FIGS. 5 and 6 may be omitted or additional steps may be present without departing from the scope of the present disclosure. Additionally, specific details regarding the components depicted in FIGS. 5 and 6 (e.g., computing device 100, first device 200, etc.) may be replaced with the content previously described with reference to FIGS. 1 to 4. .

As described above with reference to FIGS. 1 to 6 , the computing device 100 may determine the validity of the camera pose of the first device 200 using visual localization using a query image and a reference image. Additionally, the computing device 100 may re-estimate the camera pose when the camera pose becomes inaccurate (when the effective value corresponds to a predetermined condition). Accordingly, the computing device 100 may improve the accuracy of the camera pose of the first device 200 by continuously estimating the camera pose of the first device 200, which becomes inaccurate over time.

7 is a brief, general schematic diagram of an example computing environment in which embodiments of the present disclosure may be implemented.

Although the present disclosure has generally been described above as being capable of being implemented by a computing device, those skilled in the art will understand that the present disclosure can be implemented in combination with computer-executable instructions and/or other program modules that can be executed on one or more computers and/or in hardware and software. It will be well known that it can be implemented as a combination.

Typically, program modules include routines, programs, components, data structures, etc. that perform specific tasks or implement specific abstract data types. Additionally, those skilled in the art will understand that the methods of the present disclosure are applicable to single-processor or multiprocessor computer systems, minicomputers, mainframe computers, as well as personal computers, handheld computing devices, microprocessor-based or programmable consumer electronics, etc. It will be appreciated that each of these may be implemented in other computer system configurations, including those capable of operating in conjunction with one or more associated devices.

The described embodiments of the disclosure can also be practiced in distributed computing environments where certain tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

Computers typically include a variety of computer-readable media. Computer-readable media can be any medium that can be accessed by a computer, and such computer-readable media includes volatile and non-volatile media, transitory and non-transitory media, removable and non-transitory media. Includes removable media. By way of example, and not limitation, computer-readable media may include computer-readable storage media and computer-readable transmission media. Computer-readable storage media refers to volatile and non-volatile media, transient and non-transitory media, removable and non-removable, implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Includes media. Computer readable storage media may include RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital video disk (DVD) or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage. This includes, but is not limited to, a device, or any other medium that can be accessed by a computer and used to store desired information.

A computer-readable transmission medium typically implements computer-readable instructions, data structures, program modules, or other data on a modulated data signal, such as a carrier wave or other transport mechanism. Includes all information delivery media. The term modulated data signal refers to a signal in which one or more of the characteristics of the signal have been set or changed to encode information within the signal. By way of example, and not limitation, computer-readable transmission media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above are also intended to be included within the scope of computer-readable transmission media.

An example environment 1100 is shown that implements various aspects of the present disclosure, including a computer 1102, which includes a processing unit 1104, a system memory 1106, and a system bus 1108. do. System bus 1108 couples system components, including but not limited to system memory 1106, to processing unit 1104. Processing unit 1104 may be any of a variety of commercially available processors. Dual processors and other multiprocessor architectures may also be used as processing unit 1104.

System bus 1108 may be any of several types of bus structures that may further be interconnected to a memory bus, peripheral bus, and local bus using any of a variety of commercial bus architectures. System memory 1106 includes read only memory (ROM) 1110 and random access memory (RAM) 1112. The basic input/output system (BIOS) is stored in non-volatile memory 1110, such as ROM, EPROM, and EEPROM, and is a basic input/output system that helps transfer information between components within the computer 1102, such as during startup. Contains routines. RAM 1112 may also include high-speed RAM, such as static RAM, for caching data.

Computer 1102 may also include an internal hard disk drive (HDD) 1114 (e.g., EIDE, SATA)—the internal hard disk drive 1114 may also be configured for external use within a suitable chassis (not shown). Yes - a magnetic floppy disk drive (FDD) 1116 (e.g., for reading from or writing to a removable diskette 1118), and an optical disk drive 1120 (e.g., a CD-ROM for reading the disk 1122 or reading from or writing to other high-capacity optical media such as DVDs). Hard disk drive 1114, magnetic disk drive 1116, and optical disk drive 1120 are connected to system bus 1108 by hard disk drive interface 1124, magnetic disk drive interface 1126, and optical drive interface 1128, respectively. ) can be connected to. The interface 1124 for implementing an external drive includes at least one or both of Universal Serial Bus (USB) and IEEE 1394 interface technologies.

These drives and their associated computer-readable media provide non-volatile storage of data, data structures, computer-executable instructions, and the like. For computer 1102, drive and media correspond to storing any data in a suitable digital format. Although the description of computer-readable media above refers to removable optical media such as HDDs, removable magnetic disks, and CDs or DVDs, those skilled in the art will also recognize removable optical media such as zip drives, magnetic cassettes, flash memory cards, cartridges, etc. It will be appreciated that other types of computer-readable media, such as the like, may also be used in the example operating environment and that any such media may contain computer-executable instructions for performing the methods of the present disclosure.

A number of program modules may be stored in drives and RAM 1112, including an operating system 1130, one or more application programs 1132, other program modules 1134, and program data 1136. All or portions of the operating system, applications, modules and/or data may also be cached in RAM 1112. It will be appreciated that the present disclosure may be implemented on various commercially available operating systems or combinations of operating systems.

A user may enter commands and information into computer 1102 through one or more wired/wireless input devices, such as a keyboard 1138 and a pointing device such as mouse 1140. Other input devices (not shown) may include microphones, IR remote controls, joysticks, game pads, stylus pens, touch screens, etc. These and other input devices are connected to the processing unit 1104 through an input device interface 1142, which is often connected to the system bus 1108, but may also include a parallel port, an IEEE 1394 serial port, a game port, a USB port, an IR interface, It can be connected by other interfaces, etc.

A monitor 1144 or other type of display device is also connected to system bus 1108 through an interface, such as a video adapter 1146. In addition to monitor 1144, computers typically include other peripheral output devices (not shown) such as speakers, printers, etc.

Computer 1102 may operate in a networked environment using logical connections to one or more remote computers, such as remote computer(s) 1148, via wired and/or wireless communications. Remote computer(s) 1148 may be a workstation, computing device computer, router, personal computer, portable computer, microprocessor-based entertainment device, peer device, or other conventional network node, and is generally connected to computer 1102. For simplicity, only memory storage device 1150 is shown, although it includes many or all of the components described. The logical connections depicted include wired/wireless connections to a local area network (LAN) 1152 and/or a larger network, such as a wide area network (WAN) 1154. These LAN and WAN networking environments are common in offices and companies and facilitate enterprise-wide computer networks, such as intranets, all of which can be connected to a worldwide computer network, such as the Internet.

When used in a LAN networking environment, computer 1102 is connected to local network 1152 through wired and/or wireless communication network interfaces or adapters 1156. Adapter 1156 may facilitate wired or wireless communication to LAN 1152, which also includes a wireless access point installed thereon for communicating with wireless adapter 1156. When used in a WAN networking environment, the computer 1102 may include a modem 1158 or be connected to a communicating computing device on the WAN 1154 or to establish communications over the WAN 1154, such as via the Internet. Have other means. Modem 1158, which may be internal or external and a wired or wireless device, is coupled to system bus 1108 via serial port interface 1142. In a networked environment, program modules described for computer 1102, or portions thereof, may be stored in remote memory/storage device 1150. It will be appreciated that the network connections shown are exemplary and that other means of establishing a communications link between computers may be used.

Computer 1102 may be associated with any wireless device or object deployed and operating in wireless communications, such as a printer, scanner, desktop and/or portable computer, portable data assistant (PDA), communications satellite, wirelessly detectable tag. Performs actions to communicate with any device or location and telephone. This includes at least Wi-Fi and Bluetooth wireless technologies. Accordingly, communication may be a predefined structure as in a conventional network or may simply be ad hoc communication between at least two devices.

Wi-Fi (Wireless Fidelity) allows connection to the Internet, etc. without wires. Wi-Fi is a wireless technology, like cell phones, that allows these devices, such as computers, to send and receive data indoors and outdoors, anywhere within the coverage area of a base station. Wi-Fi networks use wireless technology called IEEE 802.11 (a, b, g, etc.) to provide secure, reliable, and high-speed wireless connections. Wi-Fi can be used to connect computers to each other, the Internet, and wired networks (using IEEE 802.3 or Ethernet). Wi-Fi networks can operate in the unlicensed 2.4 and 5 GHz wireless bands, for example, at data rates of 11 Mbps (802.11a) or 54 Mbps (802.11b), or in products that include both bands (dual band). .

Those skilled in the art will understand that information and signals may be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, values, symbols and chips that may be referenced in the above description include voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields. or particles, or any combination thereof.

Those skilled in the art will understand that the various illustrative logical blocks, modules, processors, means, circuits and algorithm steps described in connection with the embodiments disclosed herein may be used in electronic hardware, (for convenience) It will be understood that it may be implemented by various forms of program or design code (referred to herein as software) or a combination of both. To clearly illustrate this interoperability of hardware and software, various illustrative components, blocks, modules, circuits and steps have been described above generally with respect to their functionality. Whether this functionality is implemented as hardware or software depends on the specific application and design constraints imposed on the overall system. A person skilled in the art of this disclosure may implement the described functionality in various ways for each specific application, but such implementation decisions should not be construed as departing from the scope of this disclosure.

The various embodiments presented herein may be implemented as a method, apparatus, or article of manufacture using standard programming and/or engineering techniques. The term article of manufacture includes a computer program, carrier, or media accessible from any computer-readable storage device. For example, computer-readable storage media include magnetic storage devices (e.g., hard disks, floppy disks, magnetic strips, etc.), optical disks (e.g., CDs, DVDs, etc.), smart cards, and flash. Includes, but is not limited to, memory devices (e.g., EEPROM, cards, sticks, key drives, etc.). Additionally, various storage media presented herein include one or more devices and/or other machine-readable media for storing information.

It is to be understood that the specific order or hierarchy of steps in the processes presented is an example of illustrative approaches. It is to be understood that the specific order or hierarchy of steps in processes may be rearranged within the scope of the present disclosure, based on design priorities. The appended method claims present elements of the various steps in a sample order but are not meant to be limited to the particular order or hierarchy presented.

The description of the presented embodiments is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments without departing from the scope of the disclosure. Thus, the present disclosure is not limited to the embodiments presented herein but is to be interpreted in the broadest scope consistent with the principles and novel features presented herein.

Claims (12)

A method for visual localization performed by a computing device including a processor, comprising:
Obtaining a first query image corresponding to a first time from a first device;
Obtaining first comparison data by comparing first query feature point information of the first query image obtained using a pre-learned feature point acquisition model with reference feature point information of a pre-stored first reference image;
estimating a first camera pose of the first device based on the first comparison data;
Obtaining a second query image corresponding to a second time different from the first time from the first device;
Obtaining second comparison data by comparing second query feature point information of the second query image obtained using the pre-learned feature point acquisition model with the first query feature point information;
Estimating a second camera pose of the first device related to the second query image based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device;
Based on the first comparison data, estimating a relative camera pose between the first device and a first reference device associated with the pre-stored first reference image;
Obtaining a first query 3D estimated coordinate value of the first query image based on the relative camera pose and a first reference 3D coordinate value of the pre-stored first reference image;
Obtaining a second query 3D estimated coordinate value of the second query image based on the first query 3D estimated coordinate value and information of the inertial measurement unit;
Obtaining a second query 2D estimated coordinate value of the second query image from the second query 3D estimated coordinate value using a predetermined algorithm;
Obtaining second query 2D real coordinate values of the second query image based on the second comparison data; and
determining a value of reprojection error calculated by comparing the second query 2D estimated coordinate value and the second query 2D actual coordinate value as a valid value of at least one of the first camera pose and the second camera pose;
Including,
method.
According to claim 1,
The pre-trained feature point acquisition model is,
Learned in advance through deep learning or machine learning to obtain feature points from specific images included in specific input data,
method.
According to claim 1,
The step of acquiring the first comparison data is,
Matching reference image feature point information of each of a plurality of pre-stored reference images with the first query feature point information;
determining a matching reference image that matches the first query feature point information among the plurality of pre-stored reference images as the first reference image; and
Comparing the first query feature point information with reference feature point information of the first reference image to obtain the first comparison data;
Including,
method.
According to claim 1,
The step of estimating the first camera pose is,
estimating the first camera pose based on 2D matching coordinate values and 3D matching coordinate values of each of a plurality of matching feature points included in the first comparison data;
Including,
method.
According to claim 4,
The 2D matching coordinate value is,
Containing two-dimensional pixel coordinate values in the first query image,
method.
According to claim 4,
The 3D matching coordinate value is,
Containing coordinate values in three dimensions in a coordinate system of a first reference device associated with the pre-stored first reference image,
method.
According to claim 1,
The information of the inertial measurement unit is,
Containing information related to the movement of the first device from the first time to the second time,
method.
According to claim 1,
The step of acquiring the second query image is,
Obtaining the second query image from a second device arranged based on the first camera pose and information of the inertial measurement unit;
Including,
method.
delete According to claim 1,
If a valid value of at least one of the first camera pose or the second camera pose corresponds to a predetermined condition, re-estimating at least one of the first camera pose or the second camera pose;
Containing more,
method.
A computer program stored on a computer-readable storage medium, the computer program comprising instructions for causing a processor of a computing device for visual localization to perform the following steps, the steps comprising:
Obtaining a first query image corresponding to a first time from a first device;
Obtaining first comparison data by comparing first query feature point information of the first query image obtained using a pre-learned feature point acquisition model with reference feature point information of a pre-stored first reference image;
estimating a first camera pose of the first device based on the first comparison data;
Obtaining a second query image corresponding to a second time different from the first time from the first device;
Obtaining second comparison data by comparing second query feature point information of the second query image obtained using the pre-learned feature point acquisition model with the first query feature point information;
Estimating a second camera pose of the first device related to the second query image based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device;
Based on the first comparison data, estimating a relative camera pose between the first device and a first reference device associated with the pre-stored first reference image;
Obtaining a first query 3D estimated coordinate value of the first query image based on the relative camera pose and a first reference 3D coordinate value of the pre-stored first reference image;
Obtaining a second query 3D estimated coordinate value of the second query image based on the first query 3D estimated coordinate value and information of the inertial measurement unit;
Obtaining a second query 2D estimated coordinate value of the second query image from the second query 3D estimated coordinate value using a predetermined algorithm;
Obtaining second query 2D real coordinate values of the second query image based on the second comparison data; and
determining a value of reprojection error calculated by comparing the second query 2D estimated coordinate value and the second query 2D actual coordinate value as a valid value of at least one of the first camera pose and the second camera pose;
Including,
A computer program stored on a computer-readable storage medium.
In a computing device for visual localization,
A processor including at least one core;
a memory storing a computer program executable by the processor; and
network department;
Including,
The processor,
Obtaining a first query image corresponding to a first time from a first device,
Obtaining first comparison data by comparing first query feature point information of the first query image obtained using a pre-learned feature point acquisition model with reference feature point information of a pre-stored first reference image,
Estimating a first camera pose of the first device based on the first comparison data,
Obtaining a second query image corresponding to a second time different from the first time from the first device,
Obtaining second comparison data by comparing second query feature point information of the second query image obtained using the pre-learned feature point acquisition model with the first query feature point information,
Based on the first camera pose and information of an inertial measurement unit (IMU) included in the first device, estimate a second camera pose of the first device related to the second query image,
Based on the first comparison data, estimate a relative camera pose between the first device and a first reference device associated with the pre-stored first reference image,
Based on the relative camera pose and a first reference 3D coordinate value of the pre-stored first reference image, obtain a first query 3D estimated coordinate value of the first query image,
Obtaining a second query 3D estimated coordinate value of the second query image based on the first query 3D estimated coordinate value and the information of the inertial measurement unit,
Obtaining a second query 2D estimated coordinate value of the second query image from the second query 3D estimated coordinate value using a predetermined algorithm,
Obtaining a second query 2D real coordinate value of the second query image based on the second comparison data, and
Determining the value of the reprojection error calculated by comparing the second query 2D estimated coordinate value and the second query 2D actual coordinate value as a valid value of at least one of the first camera pose and the second camera pose,
Computing device.

Images