CN111288996A - Indoor navigation method and system based on video live-action navigation technology - Google Patents

Abstract

The invention provides an indoor navigation method based on the video real scene navigation technology. The path division nodes are obtained through modeling, and the matching field path nodes are obtained through the first-view video shooting; the path node video is processed and compressed. On the basis of , all path nodes are combined in pairs to form a video node combination pair; the shortest path corresponding to the path node combination pair is calculated to form the shortest path video data represented by the path nodes, which is the indoor navigation video data. The user obtains the location information as the starting point information according to the path node label or path node picture, selects the destination as the ending point information through the user interaction terminal, and then requests the shortest path video data that is adapted to the starting point information and the ending point information. The invention also provides an indoor navigation system. The invention overcomes the conflict of navigation accuracy and cost in the prior art, and realizes indoor navigation with no information source, high precision and low cost.

Description

Indoor navigation method and system based on video real scene navigation technology

technical field

The present invention relates to the technical field of indoor navigation, in particular, to an indoor navigation method and system based on the video real scene navigation technology.

Background technique

Indoor navigation has important practical value in real life. At present, the main indoor navigation technologies can be divided into the following three categories: indoor navigation technology based on location perception, indoor navigation technology based on signal measurement, and indoor navigation technology based on sensor, and indoor navigation technology that can be put into commercial use mostly adopts sensor-based indoor navigation technology Technology, commercial companies customize a set of base station construction solutions for specific buildings, install base stations, and complete accurate positioning and navigation inside the building through the navigation system. Although this method is mature, due to its high cost, huge amount of engineering and high maintenance cost, it is currently only used in large factories to locate goods or employees, and is limited in public places such as shopping malls, commercial institutions or hospitals. Due to the above problems, it cannot be widely used.

After searching and researching existing technologies at home and abroad, it is found that:

Indoor navigation technology is often based on indoor positioning technology, which can be divided into specific navigation types according to the type of source. At present, the research results have included WiFi signals (refer to [2] He Tao. WiFi indoor positioning algorithm based on sensor data fusion) Design and application research [D]. Zhejiang University, 2016.), Bluetooth signal, electromagnetic wave signal, including a variety of source positioning methods, the basic idea is to pre-install the signal transmitter, and realize the location information through the acceptance and return of the terminal equipment acquisition. This type of positioning technology depends to a large extent on the anti-interference and reliability of SIA, and because the base station (ie, the signal transmitter) is essential, it is particularly difficult to compress the cost of various positioning and navigation technologies. The high-end and precise signal transmitting device further increases the engineering workload for the construction of the positioning system for the power supply of the source device.

The iBeacon indoor navigation technology led by Apple has been successful and has been successfully commercialized. iBeacon is essentially a base station-based positioning and navigation technology. The iBeacon technology is mainly based on the low-power Bluetooth signal conduction technology. A large number of iBeacon signal transmitters are arranged in the target building. The user's mobile device can receive and process related signals. The position is calculated, and the indoor precise guidance is finally realized. Therefore, the promotion of iBeacon technology has the following disadvantages: First, the deployment of iBeacon base stations requires a certain amount of engineering, and the base station is powered by batteries, which makes subsequent maintenance more troublesome; due to the open protocol, iBeacon devices are easily forged to steal user mobile phone data. Therefore, Its safety is also open to question. However, thanks to Apple's influence and technical support, as well as its own technical advantages, iBeacon has been adapted and applied to a certain extent abroad. At least 30 large shopping malls and airports have adopted iBeacon technology as an indoor navigation system. Solution (refer to Wei Xiangshen. Application of iBeacon in indoor positioning system based on Microchip BM70/1 Bluetooth module [J]. World of Electronic Products, 2019, 26(02): 18-19.). However, the above technical defects still limit the development of iBeacon.

At present, with the development of other various indoor navigation and positioning technologies, many companies dedicated to the customization of indoor navigation solutions are also constantly developing their own indoor navigation systems. The more representative ones are Xi and Beidou, which mainly use TC-OFDM+BDS (refer to Zhao Xuanxuan, Han Litao, Zheng Ying, Lei Yanhui, Wu Jiayi. Review of Indoor Navigation Model Research [J]. Software Guide, 2016 , 15 (05): 1-3.) technology combination, dedicated to the application of satellite systems to indoor positioning. However, this technology mainly targets large and medium-sized projects, and uses satellite navigation systems and outdoor base stations to develop special navigation systems. Therefore, the corresponding implementation costs of this technology are relatively high. In addition to the construction of base stations, it is necessary to purchase the right to use the satellite navigation system, etc. .

In addition, there is Qianyun Map, which uses the inertial sensor in the smartphone to complete the navigation without relying on external sources. This method solves the problem of accuracy and cost very well, and the adaptation difficulty is low, but due to the characteristics of the inertial navigation itself (can be Refer to Liu Peng, Yi Feng, Zhang Ya, Wu Changcheng. Graph method for observability analysis and observable state determination of inertial navigation systems [J/OL]. Control Theory and Application: 1-9[2019-03-27].) , with the increase of navigation time, the error of inertial navigation will gradually increase, so this navigation technology is mainly used in parking lots and other places, and it is difficult to achieve the due stability in the real sense of indoor.

In addition to the above two representative companies, various types of companies rely on UWB, WiFi, LTE, BLE (refer to Sun Jiwu, Jiang Lingyun. Optimization of Opportunistic Routing Protocols in BLE Mesh Networks [J]. Journal of Nanjing University of Posts and Telecommunications (Natural Science Edition) , 2018, 38(06): 90-95.) and RFID (refer to Xiao Kailin. Indoor Navigation of Autonomous Mobile Robots Based on UHF RFID Hybrid Localization Algorithm [D]. Huazhong University of Science and Technology, 2017.) and other technology commercial indoor Navigation technologies emerge in an endless stream, providing a large number of indoor navigation solutions. Although a fairly mature industry has been formed, limited by the cost and navigation accuracy, the currently developed technologies or formed systems can be used against targets. The application scenarios are still very narrow.

To sum up, at present, various commercial navigation technologies basically rely on the existing typical indoor positioning technology, use a combination of different positioning technologies to achieve navigation, and use certain algorithms to make corrections to improve user experience. Fundamentally solve the main obstacle to the popularization of indoor navigation: the conflict between navigation accuracy and cost. Due to the technical difficulties that the cost and accuracy cannot be taken into account, the application scenarios of indoor navigation have been severely limited, so that today, when outdoor navigation shines brightly and various large buildings are erected, there is still no mature indoor navigation application available. use. How to achieve higher accuracy and lower cost has become the main obstacle to the widespread application of indoor navigation. As far as the current technical research on indoor navigation is concerned, the research idea is slightly rigid, and the too tight binding of positioning and navigation restricts the development of navigation technology. The cost of navigation adaptation and navigation accuracy have become the popularization of indoor navigation technology. How to solve this contradiction has also become an urgent problem in this field.

At present, there is no description or report of the technology similar to the present invention, and no similar materials at home and abroad have been collected.

SUMMARY OF THE INVENTION

Aiming at the above shortcomings in the prior art, the present invention provides an indoor navigation method and system based on the video real scene navigation technology. The method and system use the first-person perspective video to guide the user to complete indoor navigation, and the user only needs to follow the video perspective returned by the system to travel to the destination, which overcomes the conflict between navigation accuracy and cost in the prior art, and realizes indoor confidence-free operation. Source, high-precision, low-cost indoor navigation, laying the corresponding technical foundation for the popularization of indoor navigation.

The present invention is achieved through the following technical solutions.

According to an aspect of the present invention, there is provided an indoor navigation method based on the video real scene navigation technology, including:

On the server side:

S1, modeling the buildings that need to be adapted for navigation;

S2, on the basis of the model built in S1, divide the internal path of the building, and obtain the path division node;

S3, according to the path divided in S2, obtain the first-view video information of the corresponding building field path, and obtain the field path node adapted to the path division node;

S4, including any one or any of the following steps:

- Set the labels for each path node obtained in S3, and remove the interfering elements in the video of the actual path of the building and carry out eye-catching treatment on the iconic elements;

- Eliminate the interfering elements in the video of the actual path of the building obtained in S3 and perform eye-catching processing on the iconic elements, and extract path node pictures from all path nodes at a set time interval to form a picture library; preferably, set The predetermined time interval is 0.1s; further preferably, the information contained in each path node picture includes: the path node to which the picture belongs and the time when the picture appears in the path node.

S5, on the basis of S4, combine all the path nodes of the building in pairs to obtain all possible path node combination pairs; calculate the shortest paths corresponding to all the path node combination pairs, and form the shortest path video data represented by the path nodes , the shortest path video data is indoor navigation video data.

Preferably, in the S2, the method for dividing the internal path of the building and obtaining the path dividing node is:

Based on all the locations in the building that may be selected by the user as the navigation destination, the path between two adjacent possible destinations is used as a basic path division node, and the path between any two locations in the building is All are combined by one or more basic path division nodes.

Preferably, in said S4, the pr video editing method is used to eliminate the interfering elements in the video of the building's field path; and/or

Striking treatment of iconic elements takes a rendering approach.

Preferably, in the S5, a bidirectional A* algorithm is used to calculate the shortest path corresponding to the path node combination pair; wherein:

The bidirectional A* algorithm includes:

Take one path node in the path node combination pair as the starting point and the other path node as the end point;

Start two threads and execute two different search processes respectively, in which one thread processes the search process from the path node as the starting point to the path node as the end point, and the other thread processes the path node from the end point to the starting point The search process of the path node;

When two threads process the same path node, it is determined that a shortest path corresponding to a path node combination pair is obtained.

Preferably, the method further includes:

On the user side:

s1, the user obtains the location information as the starting point information according to the available path node labels or path node pictures, and selects the destination as the end point information through the user interaction terminal;

s2, according to the starting point information and the ending point information, the user interaction terminal requests the server side for shortest path video data adapted to the starting point information and the ending point information, and plays the video data.

Preferably, in the s1, the method for the user to obtain the location information as the starting point information according to the available path node labels is: setting the path node label in the form of a two-dimensional code, and the user obtains the location information by scanning the two-dimensional code. .

Preferably, in the s1, the method for the user to obtain the location information as the starting point information according to the available path node pictures is as follows: the user takes a picture of the path in front of him and uploads it, and the server side, according to the picture uploaded by the user, and the path node picture one by one Pixel-level comparison is performed. When the similarity with a certain path node picture reaches the set threshold, it is judged that the picture uploaded by the user is the same as the path node picture, and then the location information is obtained according to the path node picture.

Preferably, in the s2, before playing the shortest path video data, the method further includes: instructing the user to adjust the initial direction of the navigation.

Preferably, the method further includes:

On the user side:

s3, the user interaction terminal feeds back the user's evaluation result of the shortest path video data to the server side, which is used for the server side to optimize the shortest path video data; wherein:

The optimization method is: using the bidirectional A* algorithm to recalculate the shortest path corresponding to the path node combination pair.

According to a second aspect of the present invention, there is provided an indoor navigation system based on the video real scene navigation technology, characterized in that it includes:

- Set on the server side:

Building modeling module: The building modeling module performs CAD modeling for buildings that need to be navigated and adapted;

Path division module: the path division node acquisition module divides the internal path of the building on the basis of the model built by the building modeling module, and obtains the path division node;

Path node acquisition module: the path node acquisition module acquires the corresponding first-view video information of the building field path according to the path divided in the path division module, and obtains the field path node adapted to the path division node;

Video processing module: including a video processing unit, the video processing unit is used for:

Perform label setting on each path node obtained in the path node acquisition module, remove interfering elements in the video of the actual path of the building, and perform eye-catching processing on iconic elements; and/or

Eliminate the interfering elements in the video of the real path of the building obtained in the path node acquisition module, and carry out eye-catching processing on the iconic elements, and extract the path node pictures from all the path nodes at a set time interval to form a picture library;

Path calculation module: on the basis of the video processed by the video processing module, the path calculation module performs pairwise combinations of all the path nodes of the building to obtain all possible path node combinations; calculates the shortest corresponding to all the path node combinations. path to form the shortest path video data represented by path nodes;

Database module: the database module is used to save the shortest path video data obtained in the path calculation module, and/or, the picture library obtained in the video processing module;

- Set on the client side:

Path request module: the path request module obtains the user's location information as the starting point information according to the available path node labels or path node pictures, and obtains the path starting point and end point information according to the selected destination data as the end point information;

Navigation module: The navigation module retrieves the shortest path video data adapted to the path start and end point information from the database module on the server side according to the path start and end point information obtained by the path request module.

Preferably, the video processing module further comprises: a video compression unit, the video compression unit compresses the video of the building's field path processed by the video processing unit.

Preferably, in the path request module, the path node label is obtained by a position identifier that is set inside the building and corresponds to the position of the video node and corresponds to the video node label one-to-one.

Preferably, the path request module further includes: a picture matching unit, the picture matching unit performs pixel-level comparison with the path node picture one by one according to the path picture in front of the user uploaded by the user, and when the similarity with a certain path node picture reaches When the threshold is set, it is determined that the picture uploaded by the user is the same as the picture of the path node, and then the location information is obtained according to the picture of the path node.

Preferably, the navigation module further comprises: acquiring the user's evaluation result on the shortest path video data, and sending the evaluation result to the path calculation module for path optimization.

Preferably, the location identifier is set in the form of a two-dimensional code or a program code; correspondingly, the path request module includes a two-dimensional code or program code scanning unit.

Preferably, the path request module further includes an image acquisition unit for the user to acquire a picture of the path in front of him.

Compared with the prior art, the present invention has the following beneficial effects:

1. The indoor navigation method and system based on the video real scene navigation technology provided by the present invention realizes the construction of a low-cost and high-precision indoor navigation scheme through independent research and development and application of the video real scene navigation technology, and solves the cost in the prior art. It fills the gap in the existing technology and lays a technical foundation for the popularization of indoor navigation.

2. The indoor navigation method and system based on the video real scene navigation technology provided by the present invention provides a two-way A* algorithm by optimizing the A* pathfinding algorithm, which realizes the good adaptation of the A* algorithm in indoor navigation. It improves the performance and actual performance of the algorithm, and simplifies the process and workload of the actual adaptation of the video real scene navigation technology.

3. The indoor navigation method and system based on the video real scene navigation technology provided by the present invention has far-reaching significance for the commercialization and popularization of indoor navigation, which fundamentally solves the serious obstacle of the irreconcilable cost and accuracy of the traditional navigation system, It provides a new idea for indoor navigation scheme, which has great application value and practical significance.

Description of drawings

Other features, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments with reference to the following drawings:

FIG. 1 is a flowchart of an indoor navigation method based on a video real scene navigation technology provided by an embodiment of the present invention;

FIG. 2 is a schematic diagram of acquiring a visual path video node from a first perspective provided by an embodiment of the present invention;

FIG. 3 is a schematic diagram of a data flow of an indoor navigation system based on a real-time video navigation technology provided by an embodiment of the present invention;

Fig. 4 is the bidirectional A* algorithm provided by the embodiment of the present invention in a given grid map, respectively calculates 8-domain Manhattan distance, 4-domain Manhattan distance, 8-domain Euclidean distance, and 4-domain Euclidean distance operation result diagram; wherein, (a) is the running result of 8-domain Manhattan distance, (b) is the running result of 4-domain Manhattan distance, (c) is the running result of 8-domain Euclidean distance, (d) is the running result of 4-domain Euclidean distance picture.

Detailed ways

The embodiments of the present invention are described in detail below: This embodiment is implemented on the premise of the technical solution of the present invention, and provides detailed implementation modes and specific operation processes. It should be pointed out that for those skilled in the art, without departing from the concept of the present invention, several modifications and improvements can be made, which all belong to the protection scope of the present invention.

The embodiments of the present invention provide an indoor navigation method and system based on a video real scene navigation technology. The first-person perspective video is used to guide the user to complete the indoor navigation. The user only needs to follow the returned video perspective to travel to the destination. Source-free, high-precision, low-cost indoor navigation lays a corresponding technical foundation for the popularization of indoor navigation.

The technical solutions provided by the embodiments of the present invention are further described in detail below with reference to the accompanying drawings.

As shown in FIG. 1, the indoor navigation method based on the video real scene navigation technology provided by the embodiment of the present invention, on the server side, the method includes the following steps:

Step S1, CAD modeling is performed on the building that needs to be navigated and adapted;

Step S2, on the basis of the model built in step S1, the internal path of the building is divided, and the path division node is obtained; (wherein, the path division node represents the non-physical node in the division scheme of the path);

Step S3, according to the path divided in step S2, obtain the first-view video information of the corresponding building field path, and obtain the field path node adapted to the path division node; wherein, the first-view video information of the building field path is obtained. Obtained by shooting the pre-divided path from the first perspective of the camera on the spot; (wherein, the path node represents the physical node in the collected video information);

Step S4 includes any one or any of the following steps:

- Uniformly label each path node obtained in step S3, and correspond to the actual path of the building; use pr editing to remove interfering elements (such as pedestrians, etc.); do eye-catching treatment to iconic elements (such as specific store signs) , road signs, etc.);

- Eliminate interfering elements (such as pedestrians, etc.) and prominently process iconic elements (such as specific store signs, road signs, etc.) in the video of the building's field path obtained in S3, and select from all Extract path node pictures from path nodes to form a picture library;

In step S5, on the basis of the video data obtained in step S4, all path nodes of the building are combined in pairs to obtain all possible path node combinations; The shortest path video data represented by the shortest path video data is the indoor navigation video data.

Further, at the user end, the method further includes:

s1, the user obtains the location information as the starting point information according to the available path node labels or path node pictures, and selects the destination as the end point information through the user interaction terminal;

s2, according to the starting point information and the ending point information, the user interaction terminal requests the server side for shortest path video data adapted to the starting point information and the ending point information, and plays the video data.

The user walks following the perspective of the shortest path video data to complete the navigation.

Further, in the s2, before playing the shortest path video data, the method further includes: instructing the user to adjust the initial direction of the navigation.

Further, at the user end, the method further includes:

The user interaction terminal feeds back the user's evaluation result of the shortest path video data to the server side, which is used for the server side to optimize the shortest path video data; wherein, the optimization method is specifically: using a bidirectional A* algorithm, recalculating the path node combination pair the corresponding shortest path.

In this embodiment of the present invention:

The path division node is essentially a division of the path inside the building. Any path inside the building can be represented by a combination of one or more existing path nodes.

Use the first-person perspective to shoot the path nodes corresponding to the path division nodes. The path node matches the path division node. The shooting process is shown in Figure 2. What the photographer sees is the video recorded, and the user can follow the video to repeat the photographer's journey. The perspective of the video acts as a virtual guide.

The video information of the actual path of the building needs to be edited and modified, and the disturbing elements (such as pedestrians) in the video are cut out, and the iconic elements (such as key store signs) are highlighted by rendering, and finally the video can be guaranteed. Compress the video size based on the resolution to speed up the transmission rate.

The embodiment of the present invention adopts the strategy of exchanging space for time, calculates all possible paths inside the building in advance, and saves the path node sequence numbers or path node pictures corresponding to the possible paths in the database; when a user sends a path request, only It is necessary to simply match user requirements (start and end point information) and the pre-calculated path, and call the corresponding path node combination.

In the embodiment of the present invention, by setting the position identifier corresponding to the position of the path node and corresponding to the path node label one-to-one, the user can easily obtain the current location information as the starting point; or, through the path picture in front of the user uploaded by the user, Compare with the path node picture one by one at the pixel level. When the similarity with a certain path node picture reaches the set threshold, it is judged that the picture uploaded by the user is the same as the path node picture, and then the location information is obtained according to the path node picture. as a starting point.

In the embodiment of the present invention, the algorithm for pixel-level comparison with the path node pictures one by one adopts a sequential similarity detection algorithm (Sequential Similiarity Detection Algorithm, SSDA algorithm for short), and all pictures in the path node picture library are used as search pictures respectively, The user-uploaded image is used as the template image for searching. When a user-uploaded image is successfully searched in a certain image, it is considered that the user is currently in the corresponding position of the image; if there are multiple images that can be searched for the user-uploaded image, select the most The image that is close (that is, the lowest cumulative error) is regarded as the ideal image; if there is no image that meets the conditions, the user is prompted to upload it again.

In the embodiment of the present invention, the setting of the threshold value is set according to the specific terrain of the building, the video definition during implementation, and the pixel size of the picture.

The location identification can be a two-dimensional code or a small program code, and the information contained in it is the command to open the user interactive terminal and the location information of the two-dimensional code; that is, the user scans the code, and when the application is opened, the location where the user scans the code is obtained. At the same time, before playing the shortest path video data, it also includes: instructing the user to adjust the initial direction of the navigation, specifically: by loading a picture, guiding the user to rotate to adjust the body position, as the initial direction of the navigation; when the server When the terminal obtains the user's starting point information, it only knows the user's location. If the video information is played directly, the user may not know which direction to take the first step; the loaded picture can guide the user to rotate in place until the body seen If the front path is the same as the loaded image, then the direction in front of the user at this time is the initial direction of the navigation.

When the user needs to navigate, scan the QR code, the application opens and obtains the user's location. Since each location code has its unique label and location information, the server receives the user's destination, retrieves the path node combination of the origin and destination, and sends it to the user to start guiding the user's path. Alternatively, the user takes a photo of the path in front of him and uploads it, and the server compares the image uploaded by the user with the path node image one by one at the pixel level. The path node picture is the same, and then the location information is obtained according to the path node picture. The server receives the user destination, retrieves the path node combination of the starting point and the destination, and sends it to the client to start guiding the user path.

The user walks along the video perspective formed by the combination of video nodes to complete the navigation.

In the embodiment of the present invention, the bidirectional A* algorithm is used to calculate the shortest path corresponding to the combined pair of path nodes. The two-way A* algorithm used in the embodiment of the present invention is different from the traditional A* algorithm. The search method used by the traditional A* algorithm is to search from the starting point to the end point, and generate a shortest path in the process; while the two-way A* algorithm The search method adopted by the algorithm is to search from two directions at the same time, that is, one path node in the path node combination pair is used as the starting point, and the other path node is used as the end point, and the algorithm starts two threads (ie, search routes), one of which is Extends from the beginning to the end, and another thread expands from the end to the beginning. When two search routes converge to the same location, a shortest path can be obtained.

Compared with the traditional A* algorithm, the embodiment of the present invention adopts the bidirectional A* algorithm, and its running speed will be doubled, and the ideal running process is shown in FIG. 4 . In addition, the traditional A* algorithm can only obtain a fixed solution in solving complex path problems, while the bidirectional A* algorithm can obtain multiple different optimal solutions that actually exist in the process of solving some path problems. In Figure 4, (a), (b), (c), (d) are the running results of calculating the 8-domain Manhattan distance, the 4-domain Manhattan distance, the 8-domain Euclidean distance, and the 4-domain Euclidean distance using the bidirectional A* algorithm. ; In the figure, the black solid dots are the starting point and the end point, the black connecting solid line between the two black solid dots is the planned route, and the remaining diagonal black solid lines represent the node squares traversed by this plan.

Table 1 is a performance comparison table between the two-way A* algorithm and the traditional A* algorithm provided by the embodiment of the present invention; wherein M represents the Manhattan distance, E represents the Euclidean distance, and all algorithms run in the same grid map; as shown in Table 1 , you can see the comparison between the traditional A* algorithm and the bidirectional A* algorithm in a specific path problem.

Table 1

The user interaction terminal used in the embodiment of the present invention may be in the form of a WeChat applet or an APP application program.

in:

Mini program development: You can use the WeChat web developer tools (WeChat_devtools) to develop based on the JavaScript language under the Windows 10 professional operating system.

Android APP application development: You can use Android Studio to develop based on Java language under the Windows 10 professional operating system.

Server-side construction: You can install Apache and MySQL under the CentOS system, build a LAMP environment, and use the php language for development.

The embodiment of the present invention also provides an indoor navigation system based on the video real scene navigation technology, the system can be used to implement the above method, and a schematic diagram of the data flow of the system is shown in FIG. 3 .

The system includes:

- Set on the server side:

Building modeling module: CAD modeling of buildings that need to be navigated and adapted;

Path division module: On the basis of the model built by the building modeling module, the internal path of the building is divided, and the path division nodes are obtained;

Path node acquisition module: According to the path divided in the path division module, obtain the first-view video information of the corresponding building field path, and obtain the field path node adapted to the path division node;

Video processing module: set labels for each visual path point obtained in the path node acquisition module, and remove interfering elements in the video of the actual path of the building and perform eye-catching processing on iconic elements; and/or

Eliminate the interfering elements in the video of the real path of the building obtained in the path node acquisition module, and carry out eye-catching processing on the iconic elements, and extract the path node pictures from all the path nodes at a set time interval to form a picture library;

Path calculation module: On the basis of the video processed by the video processing module, all the path nodes of the building are combined in pairs to obtain all possible path node combinations; calculate the shortest paths corresponding to all the path node combinations to form the adopted path The shortest path video data represented by the node;

Database module: used to save the shortest path video data obtained in the path calculation module, and/or the picture library obtained in the video processing module;

- Set on the client side:

Path request module: obtains the path node label or path node picture where the user is located as the starting point information according to the available location identifier, and obtains the starting point and end point information of the path according to the selected destination data as the end point information;

Navigation module: According to the path start and end point information obtained by the path request module, the shortest path video data that is adapted to the path start point and end point information is retrieved from the database module on the server side.

Further, the video processing module further includes: a video compression unit for compressing the video of the actual path of the building after the interference elements are removed and the landmark elements are highlighted.

Further, in the path request module, the path node label is obtained by a position identifier that is set inside the building and corresponds to the position of the path node and corresponds to the path node label one-to-one.

Further, the path request module also includes: a picture matching unit, the picture matching unit performs pixel-level comparison with the path node picture one by one according to the path picture in front of the user uploaded by the user, and when the similarity with a certain path node picture reaches When the threshold is set, it is determined that the picture uploaded by the user is the same as the picture of the path node, and then the location information is obtained according to the picture of the path node.

Further, the location identification is set in the form of a two-dimensional code or a program code.

Correspondingly, the path request module includes a two-dimensional code or program code scanning unit.

Further, the path request module further includes an image acquisition unit for the user to acquire the path picture in front of him.

The indoor navigation method and system based on the video real scene navigation technology provided by the above-mentioned embodiments of the present invention is a brand-new indoor video real scene navigation technology; this technology is different from the traditional navigation method, and can realize low performance of external information and equipment. Dependency is a high-precision and low-cost real-time video navigation technology. When the user sends a navigation request to the server, the server will return a preset first-view path video to guide the user to complete the navigation, and the user only needs to follow the video perspective to reach the destination. During the development process, in-depth research and optimization were carried out on the division of building paths, the calculation of the shortest path between three-dimensional locations and the processing of special navigation routes to further improve the reliability of the navigation system.

The indoor navigation method and system based on the video real scene navigation technology provided by the above-mentioned embodiments of the present invention adopts the two-way A* algorithm to realize the organic combination of the algorithm, the video real scene navigation technology and the indoor path planning, and develops a video that can be well adapted to the video. Optimized indoor navigation technology for real-world navigation technology.

The indoor navigation method and system based on the video real scene navigation technology provided by the above-mentioned embodiments of the present invention use video editing software such as Pr to edit video nodes, eliminate disturbing information and add necessary eye-catching prompts to improve the user experience in practical applications; Using high-efficiency video compression technology, the video is compressed on the basis of ensuring the recognition degree, which ensures the speed of video transmission (the smoothness of the navigation video) and the user's traffic consumption (user economy).

The indoor navigation method and system based on the video real scene navigation technology provided by the above embodiments of the present invention adopts the optimized two-way A* pathfinding algorithm, which realizes the good adaptation application of the A* algorithm in indoor navigation, and improves the A* algorithm. *The performance and actual performance of the algorithm simplifies the process and workload of the actual adaptation of the video live navigation technology.

The indoor navigation method and system based on the video real scene navigation technology provided by the above-mentioned embodiments of the present invention provide two solutions for obtaining paths. Completing the fuzzy search and matching of the path; the other is through the user taking a photo of the path in front of him and uploading, using the image recognition matching and the user to determine the recognition result to double confirm the user's location as the starting point information, and complete the accurate path according to the end point information input by the user. Search for a match.

In the indoor navigation method and system based on the video real scene navigation technology provided by the above-mentioned embodiments of the present invention, path nodes are obtained through modeling, and matching field path video nodes are obtained through first-view video shooting; On the basis of processing and compression, all video nodes are combined in pairs to form a video node combination pair; the shortest path corresponding to the video node combination pair is calculated to form the shortest path video data represented by video nodes, which is the indoor navigation video data. The user obtains the location information as the starting point information according to the video node label, selects the destination as the ending point information through the user interaction terminal, and then requests the shortest path video data adapted to the starting point information and the ending point information. The indoor navigation method and system based on the video real scene navigation technology provided by the above-mentioned embodiments of the present invention overcome the conflict between navigation accuracy and cost in the prior art, and realize indoor navigation with no signal source, high precision and low cost.

Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the above-mentioned specific embodiments, and those skilled in the art can make various variations or modifications within the scope of the claims, which do not affect the essential content of the present invention.

Claims (10)

1. An indoor navigation method based on a video live-action navigation technology is characterized by comprising the following steps:

at a server side:

s1, modeling the building needing navigation adaptation;

s2, dividing the internal path of the building on the basis of the model established in S1 to obtain path division nodes;

s3, acquiring the first visual angle video information of the building real path corresponding to the path divided in S2 to obtain real path nodes matched with the path dividing nodes;

s4, removing the interference elements in the building field path video and highlighting the landmark elements, further comprising any one or more of the following steps:

-label setting each path node;

-extracting path node pictures from all path nodes at set time intervals to form a picture library;

s5, on the basis of S4, pairwise combination is carried out on all path nodes of the building, and all possible path node combination pairs are obtained; and calculating the shortest path corresponding to all the path node combination pairs to form shortest path video data represented by the path nodes, wherein the shortest path video data is indoor navigation video data.

2. The indoor navigation method based on video live-action navigation technology as claimed in claim 1, wherein in S2, the method for dividing the building internal path and obtaining the path division node comprises:

taking all places in the building which can be selected as navigation destinations by a user as a reference, taking a path between two adjacent possible destinations as a basic path dividing node, and combining the paths between any two places in the building through one or more basic path dividing nodes.

3. The indoor navigation method based on the video live-action navigation technology as claimed in claim 1, wherein in the step S4, the pr video clipping method is adopted to eliminate the interference elements in the building live-action path video; and/or

And performing eye-catching processing on the symbolic elements by adopting a rendering method.

4. The indoor navigation method based on the video live-action navigation technology of claim 1, wherein in the step S5, a bidirectional a-algorithm is adopted to calculate the shortest path corresponding to the path node combination pair; wherein:

the bidirectional a algorithm comprises:

taking one path node in the path node combination pair as a starting point and the other path node as an end point;

starting two threads, and respectively executing two incompletely identical search processes, wherein one thread processes the search process from the path node as a starting point to the path node as an end point, and the other thread processes the search process from the path node as an end point to the path node as a starting point;

when two threads process to the same path node, the shortest path corresponding to one path node combination pair is obtained by the judgment.

5. The indoor navigation method based on the video live-action navigation technology according to any one of claims 1 to 4, characterized in that the method further comprises:

at a user end:

s1, the user obtains the position information as the starting point information according to the available path node label or path node picture, and selects the destination as the end point information through the user interactive terminal;

s2, the user interaction terminal requests the server for the shortest path video data matched with the start point information and the end point information according to the start point information and the end point information, and plays the video data.

6. The indoor navigation method based on video live-action navigation technology as claimed in claim 5, wherein in s1, the method for the user to obtain the location information as the starting point information according to the available path node labels is as follows: setting the path node labels into a two-dimensional code form, and scanning the two-dimensional code by a user to acquire the position information of the user; the method for obtaining the position information as the starting point information by the user according to the available path node picture comprises the following steps: the method comprises the steps that a user takes pictures of a front path and uploads the pictures, a server side carries out pixel level comparison with path node pictures one by one according to pictures uploaded by the user, when the similarity of the server side and a certain path node picture reaches a set threshold value, the picture uploaded by the user and the path node picture are judged to be the same, and then position information is obtained according to the path node picture;

and/or

In s2, before playing the shortest path video data, the method further includes: the user is guided to adjust the initial direction of navigation.

7. The indoor navigation method based on the video live-action navigation technology according to claim 5, further comprising:

at a user end:

s3, the user interaction terminal feeds back the evaluation result of the shortest path video data by the user to the server side for the server side to optimize the shortest path video data; wherein:

the optimization method comprises the following steps: and recalculating the shortest path corresponding to the path node combination pair by adopting a bidirectional A-x algorithm.

8. An indoor navigation system based on video live-action navigation technology is characterized by comprising:

-on the server side:

a building modeling module: the building modeling module carries out CAD modeling on the building needing navigation adaptation;

a path division module: the path division node acquisition module divides the internal path of the building on the basis of the model established by the building modeling module to acquire path division nodes;

a path node acquisition module: the path node acquisition module acquires first visual angle video information of the building real path corresponding to the path node acquisition module according to the path divided by the path dividing module to obtain real path nodes matched with the path dividing nodes;

the video processing module: the video processing unit is used for removing interference elements in the building real-spot path video obtained by the path node acquisition module and performing eye-catching processing on landmark elements; and is also used for:

setting labels of all the path nodes obtained by the path node acquisition module; and/or

Extracting path node pictures from all path nodes at set time intervals to form a picture library;

a path calculation module: the path calculation module combines every two path nodes of the building on the basis of the video processed by the video processing module to obtain all possible path node combination pairs; calculating the shortest path corresponding to all path node combination pairs to form shortest path video data represented by path nodes;

a database module: the database module is used for storing the shortest path video data obtained in the path calculation module and/or a picture library obtained in the video processing module;

-at the user end:

a path request module: the path request module acquires position information of a user as starting point information according to the available path node label or path node picture, and acquires starting point and end point information of a path according to the selected destination data as end point information;

a navigation module: and the navigation module calls the shortest path video data matched with the path starting point and end point information from a database module at the server end according to the path starting point and end point information obtained by the path request module.

9. The indoor navigation system based on the video live-action navigation technology according to claim 8, further comprising any one or more of the following:

-the video processing module further comprises: the video compression unit compresses the building real-field path video processed by the video processing unit;

in the path request module, the path node labels are obtained by the position identifiers which are arranged inside the building and correspond to the positions of the path nodes, and correspond to the path node labels one by one;

-the path request module further comprises: the image matching unit compares the pre-route image uploaded by the user with route node images one by one at a pixel level, judges that the image uploaded by the user and the route node image are the same when the similarity of the pre-route image and the route node image reaches a set threshold value, and further acquires the position information according to the route node image;

-the navigation module further comprises: and obtaining an evaluation result of the user on the shortest path video data, and sending the evaluation result to a path calculation module for path optimization.

10. The indoor navigation system based on video live-action navigation technology as claimed in claim 8,

the position mark is set in the form of a two-dimensional code or a program code; correspondingly, the path request module comprises a two-dimensional code or program code scanning unit; and/or

The path request module also comprises an image acquisition unit used for acquiring the route picture before the user.

Images