Autonomous optical local positioning and navigation system

Abstract

The invention relates to an autonomous optical local positioning and navigation system comprising the following branches: a first branch for positioning according to an address and an information sign, decoding and identifying text by a technical vision device when information about a street or an advertising sign is extracted from a received video image, estimating the coordinate or position of the object according to the database; the second branch is used for positioning according to a topographic map without preprocessing; the third branch is that positioning is carried out according to the contour frame, and recognition is carried out according to contour points and line marks extracted from the original image; and a fourth branch, navigation is carried out through an optical ranging method, a final positioning solution can be obtained by combining available solutions obtained by corresponding branches, and then the final positioning solution is combined with a navigation solution in the presence of the navigation solution.

Description

Technical Field

[0001] The claimed invention relates to the field of autonomous navigation, namely, an autonomous optical local positioning and navigation system. Background Art

[0002] As a patent analogue of the invention applied for, for example, an image capturing device (see RU2463663, publication date: October 10, 2012) (1), comprising: an image forming unit for capturing a digital image; a communication unit for receiving a plurality of additional information items from a preset customized map server storing a plurality of additional information items (each additional information item corresponds to an object contained in the digital image captured by the unit); a display information extraction unit for selecting additional information to be displayed from the plurality of additional information items received by the communication unit; and an overlay display processing unit and a display unit for displaying only the additional information selected by the unit from the plurality of additional information items and overlaying it on the digital image captured by the unit.

[0003] The device (1) proposed as an analogue is not a positioning and navigation system, but a geographic information system element with video capture and image comparison capabilities for displaying and searching information.

[0004] The applicant considers that the closest analogue is the method for autonomous navigation of unmanned aerial vehicles (see RU2767477, published on March 17, 2022) (2), which consists in obtaining a reference image and a working image using an infrared optoelectronic system. The reference image is formed by solving a direct problem of radiative heat transfer based on a thermal terrain model, taking into account the incident solar radiation flux density, air temperature and flight altitude for a specified terrain segment of the flight trajectory and the corresponding astronomical period. The working image is recorded during nadir photography using the infrared optoelectronic system. The obtained reference and working infrared images should be thresholded to form reference and working matrices of the object's reference points (contours). The maximum two-dimensional matrix of the cross-correlation function is calculated and found, and the geographical location of the maximum value of the cross-correlation function is estimated using the reference matrix of the reference points (contours). The unmanned aerial vehicle is autonomously navigated using this estimate of the position of the maximum value of the cross-correlation function based on the measured values ​​of the heading, inclination, pitch and altitude of the unmanned aerial vehicle. The technical result is an improved accuracy in determining the navigation parameters of the unmanned aerial vehicle in the event of incorrect operation of the global satellite navigation system receiver.

[0005] The closest analogue (2) describes an autonomous optical navigation system for unmanned aerial vehicles using an infrared optoelectronic system. In this case, filming is performed in the infrared range along the lower surface (the camera is aimed at the lowest point).

[0006] The proposed system differs in this respect in that the initial measurements are made in the optical range, which has lower contrast and more contour points and color variations compared to infrared (thermal imaging). Optical cameras are mounted around the photosensitive matrix of the corresponding band, which, unlike infrared cameras, form different information flows. Therefore, the same terrain captured by cameras of different bands will form different pictures, requiring different reference maps and different processing.

[0007] The next difference is the aiming direction of the camera. In the system of patent (2), the aiming direction is aiming at the nadir, i.e. strictly downwards towards the lower surface. In the proposed system, the scanning of the camera is omnidirectional - a circular horizontal scan when moving on the surface of the earth, and when moving in the air, negative pitch aiming and large aperture scanning are added. This allows the use of three-dimensional maps and the processing of street information signs with technical vision and text recognition methods through neural networks or related algorithms, which is not envisaged in patent (2). Unlike optical scanning systems, video capture of information signs using thermal imaging or infrared cameras is not possible.

[0008] The next difference lies in the mapping basis of the system. The system of patent (2) utilizes a two-dimensional thermal projection of the underlying surface image in the form of discrete threshold contours along the horizontal plane of the thermal trace. The proposed system shall use panoramic optical measurements of streets, terrain and other objects in a three-dimensional map with contour lines and points. The map may include a database of local information signs such as road, street or advertising signs.

[0009] The next difference lies in the positioning method. The patented (2) system solves the navigation problem in a local Cartesian coordinate system. In the proposed system, referencing the local system is an intermediate step, and the street address is finally referenced. Summary of the invention

[0010] The technical result of the invention applied for is to improve the versatility of the autonomous optical local positioning and navigation system.

[0011] The technical achievement applied for is achieved by an autonomous optical local positioning and navigation system, whose algorithm is implemented by separate positioning and navigation branches. In the first branch, positioning is performed based on addresses and information signs. When information about street or advertising signs is extracted from the received video images, the text is decoded and recognized by technical visual equipment, and the coordinates or position of the object is estimated based on the database. In the second positioning branch, the target task is solved by topographic measurement without preprocessing. In the third positioning branch, the task is solved by contour frames, where recognition is performed based on contour points and line marks extracted from the original image. In the fourth branch, the navigation task is introduced using optical ranging, and the final positioning solution can be obtained by merging the available solutions obtained by the corresponding branches, and then merged with the navigation solution if there is a navigation solution.

[0012] In one embodiment, the system includes a camera that generates a digital video stream, which is provided to a primary traffic processing module via a high-speed channel. The primary traffic processing module receives the information stream from the camera and decodes discrete frames of the video image. The discrete frames are transmitted to multiple modules for processing: a video frame comparison module, an information marker search module, a comparison contour primary filtering module, and an external navigation sensor interface unit.

[0013] In one embodiment, the system includes an information sign search module for searching for information and advertising signs, address signs, and other identifiers that can locate an area where an object is located or determine its exact location.

[0014] In one embodiment, the system includes an address filtering module for identifying text information and advertising signs, address information through correlation analysis or neural network technology, and then comparing it with the address information in the information address database.

[0015] In one embodiment, the system includes a video frame comparison module, which is used to perform positioning by analyzing and comparing video frames from a system camera and corresponding frames from a local image library. The input parameters of the module are the captured frames of the camera and the reference images of the database.

[0016] In one embodiment, the system includes a contrast contour primary filtering module for preparing an image for processing.

[0017] In one embodiment, the system includes a contour frame comparison module - a mathematical positioning device that estimates the position of an object by comparing the video image contour line map obtained from the comparison contour primary filtering module with an image from a database, and at the output of this module, a calculated position is obtained based on a statistical assumption that the reference image and the observed image contour coincide.

[0018] In one embodiment, the system includes an optical ranging module, by which a navigation solution can be achieved, and coordinates and speed can be estimated by observing the displacement of a contour image with relevant analysis elements, and the input end of the module receives an image that has undergone primary filtering of contrast contours, thereby significantly reducing the computational power load and the size of the memory used.

[0019] In one embodiment, the system includes an obstacle localization module for detecting and mapping the outlines of obstacles on the object path that hinder the execution of the planned route, and providing outline images from a camera at the module input end, in which correlation analysis or artificial intelligence methods are used to identify the overall obstacles and locate their positions.

[0020] In one embodiment, the system includes an instantaneous solution and ranging solution integration module, which is used to obtain an optimal weighted estimate of the autonomous optical system as a solution to a linear optimization problem, and provides a statistically weighted estimate of the position of the object of the relative solution of the ranging solution and the instantaneous solution at the input.

[0021] In one embodiment, the system comprises a final navigation processing module for statistically combining the final solution of the optical system and the measurements of the external navigation sensors, if they are available, and a database image positioning and optimization search module for optimizing the search for reference images in a database, in which case the information address database contains local addresses and textual text index information related to a geographical basis. The local image library contains panoramic index images of streets, buildings, monuments and other visual reference points related to a geographical basis. This information is used for the operation of the comparison module, and the local electronic contour database contains contour lines and points with three-dimensional references in a geographical coordinate system, which reproduce the contrast transitions of the topographic photographic images and are used to store reference contours used by the contour frame comparison module.

[0022] In one embodiment, the system includes a temporal solution correlation module for obtaining a comprehensive solution of all positioning modules.

[0023] In one embodiment, the system can be integrated with autopilots of automated ground vehicles, with navigation and piloting complexes of unmanned and manned aerial vehicles designed for airplanes or helicopters (aerial drones), and with open navigation systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The essence of the present invention is illustrated by the accompanying drawings, in which:

[0025] Figure 1 shows the operating cycle of the autonomous optical local positioning and navigation system in navigation mode;

[0026] Figure 2 The algorithm logic diagram of the autonomous optical local positioning and navigation system in navigation mode is shown.

[0027] Figure 1 : A – shooting and capturing of images, Б – receiving measurements from connected sensors (if any), B – processing of optical and measurement information by the system, Γ – providing geolocation information to the user.

[0028] 1- Camera;

[0029] 2-Primary traffic processing module;

[0030] 3-Information sign search module;

[0031] 4- Video frame comparison module;

[0032] 5- Contrast profile primary filtering module;

[0033] 6- External navigation sensor interface unit (inertial navigation system, global satellite navigation system receiver, magnetometer, etc.);

[0034] 7-Address filtering module;

[0035] 8-contour frame comparison module;

[0036] 9- Optical ranging module;

[0037] 10-Obstacle positioning module;

[0038] 11- Instantaneous deallocation module;

[0039] 12-Instantaneous solution and ranging solution integration module;

[0040] 13- Final navigation processing module;

[0041] 14- Database image positioning and optimization search module;

[0042] 15- Information address database;

[0043] 16-Local image library;

[0044] 17- Local electronic profile library;

[0045] 18-Human-machine interface: input and output; navigation information display; route planning; mode control.

[0046] These figures do not include, nor limit, all implementations of the present technology solution, but are merely illustrative material of one particular embodiment thereof. DETAILED DESCRIPTION

[0047] The following will introduce an implementation of the autonomous optical local positioning and navigation system, which does not limit all implementations of the claimed invention.

[0048] The autonomy of a positioning system refers to the ability of the system to determine the position, spatial orientation and speed of movement of an object, as well as the direction and route of further movement relative to a given map basis, without the use of external measurement sources or information support during navigation determination.

[0049] In modern navigation equipment for large-scale everyday transport and industrial transport, external navigation measurement sources are dedicated radio distance measurement systems (and angle measurement systems in individual cases) of external operators: satellite navigation (GPS / GLONASS, etc.), mobile communication networks or local wireless networks in triangulation mode, pulse phase radio navigation systems and other infrastructure that provide navigation services locally or globally.

[0050] External information support systems are geolocation services using the Internet or other information support systems, providing real-time or quasi-real-time calculated positioning data upon request from the user. Autonomous navigation and positioning systems mean solving the target task without using the above systems, as well as without using any promising navigation service developments of external suppliers during their use for the intended purpose.

[0051] Local positioning refers to the ability of the system to make navigation decisions within a given local area, relative to a given geodetic or cartographic base, limited by the system's ability to store and process uploaded geographic data. Regional limitations are related to the maximum capacity of the geographic database carrier, processing speed, and the availability of data pre-uploaded to a separate computer.

[0052] Thus, all operations for determining the geolocation of an object are carried out by the autonomous technical equipment of the system, without the use of signals from external systems in basic mode. The main method of geolocation of the autonomous system is to calculate geolocation data in real time from the video images from the camera, by comparing the visible image with a catalog of preloaded geographical data in the local database of the system's computing device. The system can be built on the basis of a computing device of a smartphone, tablet, industrial or personal computer, or a vehicle built-in computer, as well as special devices - video recorders, rangefinders and navigators. Depending on the type of computer, the localization task is solved by technical vision, using correlation processing methods or neural network solving. The camera can be either a combined (built-in) camera with computing equipment or an external camera connected to a high-speed interface.

[0053] General principles and operating procedures of an autonomous optical local positioning and navigation system

[0054] 1. System operation preparation. In this phase, the functionality of the equipment is tested and the geographic database is loaded into the system computer. The database is a set of ordered geometric elements with cartographic or geodetic references that describe contour lines, topographic points and visual reference points, sufficient to provide unambiguous positioning in the local area, with redundancy to ensure solution stability.

[0055] 2. Instantaneous geolocation mode. Instantaneous geolocation is used at device startup or system startup to obtain an instantaneous solution and is achieved by a comprehensive search of geographic database images and comparison with observed video images, or by user input of data from external sources or adjacent navigation systems. At the time of initial association, the target task - calculating the coordinates and orientation of the object - can be solved instantaneously.

[0056] 3. Navigation calculation mode. When performing navigation calculations, digital distance measurement (based on video image shearing) is combined with geolocation determination using correlation or neural network video image comparison methods. Switch from instantaneous geolocation mode to navigation mode. When using the distance measurement navigation method, the computer load is significantly reduced and the speed parameters for object navigation can be calculated.

[0057] 4. Smooth the solution. This step can be selected if the optical measurement channel is integrated with the system's inertial or magnetic sensors (if available). Using inertial and magnetic sensors can speed up technical vision work and achieve higher navigation accuracy.

[0058] 5. Additional functions of the system. Additional functions include setting and recalculating routes, locating obstacles and detecting hazards in the path of objects, and notifying detected obstacles using voice prompts or visual and text indications.

[0059] Figure 1 Shows the system operating cycle in navigation mode

[0060] Figure 1 ——System working cycle in navigation mode.

[0061] System operation algorithm

[0062] The system runs the algorithm to execute the work cycle steps. Its main steps are described in the previous paragraph and constitute the standard steps:

[0063] Get the measured value;

[0064] ·deal with;

[0065] Display the results.

[0066] Therefore, the system should include corresponding measuring, processing and display devices. The measurement source is an optical digital camera with a digital deterministic interface. Processing is performed offline by a computing device using the system's functional software. The results are displayed on the processing device or on the screen of a connected monitor using a geographic information system digital map.

[0067] The overall logic of the system navigation and positioning is as follows: the preprocessed video image enters three computational positioning branches and one navigation branch. At the same time, depending on the hardware platform and target use, as well as the availability of the source database, some branches of the algorithm may be closed or not involved in solving the target task.

[0068] The first branch - positioning based on addresses and information signs, when street or advertising sign information is extracted from the received video image, the text is decoded and recognized through technical visual equipment, and the coordinates or position of the object is estimated based on the database. In this mode, positioning is performed based on a single shot of the surrounding environment or point by point in sequence.

[0069] Second branch - positioning based on terrain photography without preprocessing. Important reference points are panoramas of streets and industrial facilities, natural landscapes, monuments, architectural facilities and other objects with unique image features. These images are processed by visual algorithms, especially neural networks, by identification and comparison with the corresponding database.

[0070] The third branch - localization by contour frame, is similar to the second branch in meaning, except that the recognition is not performed on the basis of the original image, but on the basis of contour points and line marks extracted from the original image. The original image is converted into a set of geometric primitives, which are compared with similar images of the corresponding database by correlation or neural network machine vision methods. The difference between this method and the method implemented by the original image is that this method is more efficient, requires a smaller source database and provides higher localization accuracy.

[0071] The fourth branch is navigation by optical ranging. This method is a recursive navigation algorithm based on the displacement of observation points. It does not require an image database, but requires an initial position reference, which can be obtained from a positioning solution or an external data source.

[0072] The final positioning solution can be obtained by merging the available solutions obtained by the corresponding branches, and then merged with the navigation solution if there is a navigation solution. Based on the obtained solution, the route is set and clarified taking into account the detected obstacles. It is then displayed on the user interface or voice notification is made. When using the system as a personal assistant for disabled people, especially those with visual impairments, the personal assistant instructions integrated with the system are issued at a stepped frequency and synthesized according to the user's ability limitations.

[0073] The main source of online measurement information is the camera (1) in the equipment composition, such as Figure 2 As shown. The camera (1) forms a digital video stream, which enters the primary traffic processing module (2) through a high-speed channel. Depending on the working mode of the camera, the stream either receives a single-frame video image or a complete continuous dynamic video stream with an automatic sequential frame switching function.

[0074] The primary processing module (2) receives the information stream from the camera and decodes the discrete frames of the video image. These frames are sent to multiple modules for processing: the video frame comparison module (4), the information mark search module (3) and the contrast contour primary filtering module (5). Depending on the working mode, not all of the above modules may be involved.

[0075] There are several main modes of navigation positioning:

[0076] Instantaneous positioning mode. In this mode, the position of an object is determined based on an instantaneous image of the surrounding terrain. The instantaneous solution is obtained by the instantaneous solution association module (11);

[0077] Sequential positioning mode. In this mode, instantaneous positioning tasks are solved sequentially according to the user's requirements or a given schedule;

[0078] Navigation calculation mode. In this mode, the main calculation module is the optical ranging module (9). This mode is only possible in the case of continuous video recording. The ranging calculation requires an initial navigation approximation, which is obtained either from the instantaneous positioning mode or from an external navigation sensor.

[0079] The information sign search module (3) is used to search for information and advertising signs, address signs and other identifiers that can locate the area where the object is located or determine its exact location. This module can only be enabled if there is an information address database and only exists if the system implements the information sign navigation mode. This module does not directly read the address information, but only locates its possible location in the image. The detected address information signs enter the address filtering module (7).

[0080] The address filtering module (7) is used to identify text information, advertising signs and address information through correlation analysis or neural network technology, and then compare them with the address information in the information address database.

[0081] The video frame comparison module (8) is used to solve the positioning task by analyzing and comparing the video frames from the system camera and the corresponding frames from the local image library. The comparison is performed using technical visual methods and neural network technology. The result of the comparison is a statistical estimate of the image matching hypothesis. If qualified, the geometric calculation of the object position is performed. The input parameters of the module are the camera frame and the reference image of the database.

[0082] The contour primary filtering module (5) is used to prepare the image to be processed. Any image contains contrast contour lines and points, which form a unique combination of captured frame pictures. Primary filtering eliminates the redundancy of processed information and reduces the amount of allocated data, thereby greatly improving the response speed and increasing the capacity of the database.

[0083] Contour frame comparison module (8) - a mathematical positioning device that estimates the position of an object by comparing the contour line map of the video image obtained from the Contour Comparison Primary Filter module (5) with images from the database. The algorithm is implemented either by correlation or by a neural network solution, depending on the capabilities of the hardware and system infrastructure. At the output of this module, the calculated position is obtained based on a statistical assumption of the coincidence of the reference image and the observed image contours.

[0084] Optical ranging module (9) - implements the navigation solution and estimates the coordinates and velocity by displacement of the observed profile image with relevant analysis elements. The module receives at its input an image that has been primary filtered by contrast profiles, which significantly reduces the computational power load and the size of the memory used. The optical ranging works in a recursive loop requiring an initial approximation and a continuous supply of input profile images. In the event of a loop pause, the algorithm requires reinitialization of the initial approximation at the next start. The initial approximation can be a solution from the positioning module or an external source, including manually entered coordinates and azimuths.

[0085] The obstacle localization module (10) is used to detect and draw the outline of obstacles on the object path that hinder the execution of the planned route. The outline image from the camera is provided at the module input. Correlation analysis or artificial intelligence methods are used in these images to identify the overall obstacles and locate their positions. Then, when monitoring the execution of the route, the obstacles are bypassed and recalculated.

[0086] The instantaneous solution correlation module (11) is used to obtain the comprehensive solution of all positioning modules. Depending on the system configuration, the composition of the positioning modules may be different. If there is more than one module, these solutions need to be merged into a combined solution. The combined solution is performed by the simplest filter.

[0087] External navigation sensor interface unit (6). The system is not a closed algorithm and hardware structure and can be expanded by adding connections to external navigation sensors that are not system structural elements. These interfaces provide information exchange that can improve the conditionality of the task by increasing data redundancy, which can be used both to improve the accuracy and reliability of the integrated solution and for initial approximation and statistical control of the integrity of the measurement and solution.

[0088] The instantaneous solution and ranging solution integration module (12) is used to obtain the optimal weighted estimate of the autonomous optical system as a solution to the linear optimization problem. A statistical weighted estimate of the position of the relevant solution object of the ranging solution and the instantaneous solution is provided at the input. This can improve the accuracy and reliability of the navigation solution while reducing the probability of sudden increase and decrease of the solution.

[0089] The final navigation processing module (13) is used to statistically combine the final solution of the optical system and the measurements of the external navigation sensors, if they are available. This processing can clarify and smooth the solution, avoid solution explosion, and improve the stability of the solution with respect to unmodeled, unpredictable distortions and bad initial conditions.

[0090] The database image positioning and optimization search module (14) is used to optimize the search for reference images in the database. Considering the workload of the image optical processing algorithm, the comprehensive search and subsequent processing of reference information will cause a large delay in the processing cycle. This module searches by position prediction and expands the search range of the database to detect reference images.

[0091] The information address database (15) contains local addresses and text index information related to the geographical basis. Based on these texts, the system can roughly locate according to street signs and advertising texts. This database is designed for the operation of the address filtering module.

[0092] The local image library (16) contains panoramic index images of streets, buildings, monuments and other visual reference points related to the geographical basis. This information is designed to be used in the operation of the video frame comparison module.

[0093] The local electronic contour database (17) contains contour lines and points with three-dimensional references in a geographic coordinate system, which reproduce the contrast transitions of the topographic photographic image. This library is used to store reference contours used by the contour frame comparison module.

[0094] The human-machine interface (18) is a graphical appearance for mode management, preparing the system for operation, and displaying and notifying the user of the current navigation situation. Mode management includes selecting a positioning method and pairing a device, setting a route, and controlling the display and voice assistant operation. System operation preparation includes uploading map information, checking data integrity, and preparing the entire product for operation. Notifying the user of the current navigation situation can be done in a graphical mode, showing the map basis, the set route, or using a voice assistant to notify the user of the current location, and notifying the direction of movement along the route and possible obstacles or special situations along the way.

[0095] Industrial Applications

[0096] The scope of application of the autonomous optical local positioning and navigation system is personal and traffic navigation when moving in urban built-up areas, urbanized areas, terrain areas with visual landmarks, etc. A separate application is personal electronic assistants for people with disabilities, including visual impairments, when the system provides not only object positioning, but also warnings of obstacles on the way. The system can be integrated both with the autopilot of automated ground vehicles and with future navigation and piloting complexes of unmanned and manned aerial vehicles designed with airplanes or helicopters (drones for aerial photography), and with open navigation systems, forming an additional flow of navigation and situation information.

Claims

1. An autonomous optical local positioning and navigation system, comprising the following branches: a first branch, positioning based on addresses and information signs, when information about streets or advertising signs is extracted from received video images, the text is decoded and recognized by technical visual equipment, and the coordinates or position of the object is estimated based on the database; a second branch, positioning based on topographic maps without preprocessing; a third branch, positioning based on contour frames, where recognition is performed based on contour points and line marks extracted from the original image; a fourth branch, navigation through optical ranging, the final positioning solution is obtained by merging the available solutions obtained by the corresponding branches, and then merged with the navigation solution if there is a navigation solution.

2. The autonomous optical system according to claim 1, characterized in that It includes a camera that generates a digital video stream, which is provided to a primary traffic processing module through a high-speed channel. The primary traffic processing module receives the information stream from the camera and decodes discrete frames of the video image. The discrete frames are transmitted to multiple modules for processing: a video frame comparison module, an information mark search module, a comparison loop primary filtering module and an external navigation sensor interface unit.

3. The autonomous optical system according to claim 1 or 2, characterized in that Contains an information sign search module that searches for information and advertising signs, address signs, and other identifiers that can locate the area where an object is located or determine its exact location.

4. The autonomous optical system according to claim 1 or 2, characterized in that Contains an address filtering module, which is used to identify text information and advertising signs, address information through correlation analysis or neural network technology, and then compare it with the address information of the information address database.

5. The autonomous optical system according to claim 1 or 2, characterized in that It includes a video frame comparison module, which is used to locate by analyzing and comparing the video frames from the system camera and the corresponding frames from the local image library. The input parameters of this module are the camera's shooting frames and the reference images of the database.

6. The autonomous optical system according to claim 1 or 2, characterized in that Contains a contrast contour primary filtering module that prepares the image for processing.

7. The autonomous optical system according to claim 1 or 2, characterized in that It contains a contour frame comparison module - a mathematical positioning device, which estimates the position of the object by comparing the contour line map of the video image obtained from the comparison contour primary filtering module with the image from the database. At the output of this module, the calculated position is obtained based on the statistical assumption estimation of the coincidence of the reference image and the observed image contour.

8. The autonomous optical system according to claim 1 or 2, characterized in that It contains an optical ranging module, through which the navigation solution is realized, and the coordinates and speed are estimated by the displacement of the observed profile image with relevant analysis elements. The module input receives the image after primary filtering of the contrast profile, which significantly reduces the computing power load and the size of the memory used.

9. The autonomous optical system according to claim 1 or 2, characterized in that It includes an obstacle localization module, which is used to detect and draw the outline of obstacles on the object path that hinder the execution of the planned route. The outline image from the camera is provided at the input of the module. Correlation analysis or artificial intelligence methods are used in these images to identify the overall obstacles and locate their positions.

10. The autonomous optical system according to claim 1 or 2, characterized in that It includes an instantaneous solution and ranging solution integration module, which is used to obtain the optimal weighted estimate of the autonomous optical system as a solution to a linear optimization problem, and provides a statistically weighted estimate of the position of the object of the ranging solution and the instantaneous solution at the input.

11. The autonomous optical system according to claim 1 or 2, characterized in that It contains a final navigation processing module for statistical joint processing of the final solution measured by the optical system and the external navigation sensors, provided that they are available, and a database image positioning and optimization search module for optimizing the search for reference images in the database, in this case, the information address database contains local addresses and text text index information related to the geographical basis, the local image library contains panoramic index images of streets, buildings, monuments and other visual reference points related to the geographical basis, which information is used for the operation of the comparison module, and the local electronic contour database contains contour lines and points with three-dimensional reference in the geographical coordinate system, which reproduce the contrast transition of the topographic photographic image and is used to store reference contours used by the contour frame comparison module.

12. An autonomous optical system according to any one of the preceding claims, characterized in that Contains the instantaneous solution correlation module, which is used to obtain the comprehensive solution of all positioning modules.

13. An autonomous optical system according to any one of the preceding claims, characterized in that The system can be integrated with autopilots of automated ground vehicles, with navigation and piloting complexes of unmanned and manned aerial vehicles designed for aircraft or helicopters (aerial drones), as well as with open navigation systems.

Images