CN110765224A

CN110765224A - Processing method of electronic map, vehicle vision repositioning method and vehicle-mounted equipment

Info

Publication number: CN110765224A
Application number: CN201911021579.9A
Authority: CN
Inventors: 何潇; 熊坤; 张丹
Original assignee: Uisee Technologies Beijing Co Ltd
Current assignee: Uisee Technologies Beijing Co Ltd
Priority date: 2019-10-25
Filing date: 2019-10-25
Publication date: 2020-02-07

Abstract

The embodiment of the disclosure relates to a processing method of an electronic map, a vehicle vision repositioning method and vehicle-mounted equipment, wherein the vehicle vision repositioning method is used for a special scene, the special scene has character semantic uniqueness, and the method comprises the following steps: acquiring an image acquired by a visual sensor; identifying characters in the image; matching the characters with characters inserted in advance in the electronic map; and performing relocation based on the matching result. In the embodiment of the disclosure, for a scene with character semantic uniqueness, the characters in the scene image are identified, and the characters and global coordinates of a plurality of key features are inserted into an electronic map, so that subsequent vehicle relocation is facilitated.

Description

Processing method of electronic map, vehicle vision repositioning method and vehicle-mounted equipment

Technical Field

The disclosed embodiments relate to the technical field, and in particular, to a processing method of an electronic map, a method for vehicle vision repositioning, a vehicle-mounted device, and a non-transitory computer-readable storage medium.

Background

The visual repositioning method is a method for placing a vehicle at any unknown position in a scene, and obtaining the current positioning of the vehicle only through a map and current frame image information without depending on the image information of front and rear frames.

At present, the visual repositioning method mainly comprises a characteristic point method and a direct method, and a method for enhancing such as posting a two-dimensional code by utilizing artificial patterns. The former two methods mainly use the low-level features or pixel information in the image to code and describe the whole image, and then match with the key frame code stored in the map to realize repositioning.

However, since the first two methods only use low-level features of an image, the low-level features are features that can be automatically extracted from the image without any shape or spatial relationship information, and for example, a common threshold method is a low-level feature extraction method as point processing. The description of the low-level features to the image is not intuitive, the distinguishing power to different scenes is not strong, the situation of mismatching is easy to occur, meanwhile, the robustness to scene changes such as illumination changes and the like is poor, in addition, the former two methods need to perform traversal matching to the global map during relocation, and the time efficiency is not high; the latter method needs to manually paste the enhanced patterns, which increases the labor cost and causes certain damage to the original scene, and the traditional enhanced pattern processing algorithm has low adaptability to illumination changes.

The above description of the discovery process of the problems is only for the purpose of aiding understanding of the technical solutions of the present disclosure, and does not represent an admission that the above is prior art.

Disclosure of Invention

To solve at least one problem of the prior art, at least one embodiment of the present disclosure provides a processing method of an electronic map, a method of vehicle vision repositioning, an on-board device, and a non-transitory computer-readable storage medium.

In a first aspect, an embodiment of the present disclosure provides a method for processing an electronic map, where the method includes:

acquiring images and vehicle positioning information acquired by a vision sensor;

identifying characters in the image;

determining global coordinates of a plurality of key features corresponding to the characters based on the vehicle positioning information;

inserting the global coordinates of the character and the plurality of key features into an electronic map.

In a second aspect, the disclosed embodiment further provides a method for vehicle vision repositioning, which is used for a special scene, where the special scene has character semantic uniqueness, and the method includes:

acquiring an image acquired by a visual sensor;

identifying characters in the image;

matching the characters with characters inserted in advance in the electronic map;

and performing relocation based on the matching result.

In a third aspect, an embodiment of the present disclosure further provides an on-board device, including: a processor and a memory; the processor is adapted to perform the steps of the method according to the first or second aspect by calling a program or instructions stored in the memory.

In a fourth aspect, the disclosed embodiments also propose a non-transitory computer-readable storage medium for storing a program or instructions for causing a computer to perform the steps of the method according to the first or second aspect.

It can be seen that in at least one embodiment of the present disclosure, for a scene with semantic uniqueness of characters, the characters in the scene image are identified, and the characters and global coordinates of a plurality of key features are inserted into the electronic map, so as to facilitate subsequent vehicle relocation.

Drawings

To more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art according to the drawings.

Fig. 1 is an overall architecture diagram of an intelligent driving vehicle provided by an embodiment of the present disclosure;

FIG. 2 is a block diagram of an intelligent driving system provided by an embodiment of the present disclosure;

FIG. 3 is a block diagram of another intelligent driving system provided by embodiments of the present disclosure;

FIG. 4 is a block diagram of a map processing module provided by an embodiment of the present disclosure;

FIG. 5 is a block diagram of a relocation module provided by an embodiment of the present disclosure;

FIG. 6 is a block diagram of an in-vehicle device provided by an embodiment of the present disclosure;

fig. 7 is a flowchart of a processing method of an electronic map according to an embodiment of the present disclosure;

FIG. 8 is a flowchart of a method of vehicle vision repositioning provided by an embodiment of the present disclosure;

FIG. 9 is a schematic view of a field of a warehouse provided by an embodiment of the present disclosure;

fig. 10 is a schematic view of an urban road scene provided in an embodiment of the present disclosure.

Detailed Description

In order that the above objects, features and advantages of the present disclosure can be more clearly understood, the present disclosure will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the embodiments described are only a few embodiments of the present disclosure, and not all embodiments. The specific embodiments described herein are merely illustrative of the disclosure and are not intended to be limiting. All other embodiments derived by one of ordinary skill in the art from the described embodiments of the disclosure are intended to be within the scope of the disclosure.

It is noted that, in this document, relational terms such as "first" and "second," and the like, may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.

Aiming at the problems that the existing visual repositioning method (a characteristic point method and a direct method) only utilizes low-level information of an image, the description of the image is not intuitive, the distinguishing power of different scenes is not strong, and mismatching is easy to occur, the processing scheme of the electronic map disclosed by the embodiment of the invention is used for facilitating subsequent vehicle repositioning by identifying characters in a scene image and inserting the characters and global coordinates of a plurality of key features into the electronic map for scenes with character semantic uniqueness. The key features may be points, lines, rectangles, and other geometric shapes. Furthermore, the embodiment of the disclosure also provides a vehicle vision repositioning scheme, in the vehicle repositioning process, the repositioning is realized by identifying characters in the image and matching the pre-inserted characters, and the characters with semantic uniqueness in the scene are utilized, so that the visual positioning characteristic is realized, and the condition of mismatching is reduced.

In some embodiments, the processing scheme of the electronic map and the scheme of the vehicle vision repositioning provided by the embodiment of the disclosure can be applied to intelligent driving vehicles. Fig. 1 is an overall architecture diagram of an intelligent driving vehicle according to an embodiment of the present disclosure.

As shown in fig. 1, the smart driving vehicle includes: sensor groups, smart driving system 100, vehicle floor management systems, and other components that may be used to propel a vehicle and control the operation of the vehicle.

And the sensor group is used for acquiring data of the external environment of the vehicle and detecting position data of the vehicle. The sensor group includes, for example, but not limited to, at least one of a camera, a laser radar, a millimeter wave radar, an ultrasonic radar, a GPS (Global positioning system), and an IMU (Inertial Measurement Unit).

In some embodiments, the sensor group is further used for collecting dynamic data of the vehicle, and the sensor group further includes, for example and without limitation, at least one of a wheel speed sensor, a speed sensor, an acceleration sensor, a steering wheel angle sensor, and a front wheel angle sensor.

The intelligent driving system 100 is used for acquiring data of a sensor group, and all sensors in the sensor group transmit data at a high frequency in the driving process of the intelligent driving vehicle.

The intelligent driving system 100 is further configured to perform environment sensing and vehicle positioning based on the data of the sensor group, perform path planning and decision making based on the environment sensing information and the vehicle positioning information, and generate a vehicle control instruction based on the planned path, so as to control the vehicle to travel according to the planned path.

In some embodiments, the intelligent driving system 100 is further configured to obtain images and vehicle positioning information collected by the vision sensor; identifying characters in the image; determining global coordinates of a plurality of key features corresponding to the characters based on the vehicle positioning information; global coordinates of the character and the plurality of key features are inserted into the electronic map.

In some embodiments, the intelligent driving system 100 is further configured to acquire an image captured by a visual sensor in a special scene; identifying characters in the image; matching the characters with characters inserted in advance in the electronic map; and performing relocation based on the matching result. The special scene has character semantic uniqueness, and the special scene is, for example: parking lots, city roads with signboards, highways, and the like.

In some embodiments, the smart driving system 100 may be a software system, a hardware system, or a combination of software and hardware. For example, the smart driving system 100 is a software system running on an operating system, and the in-vehicle hardware system is a hardware system supporting the operating system.

In some embodiments, the smart driving system 100 is further configured to wirelessly communicate with a cloud server to interact with various information. In some embodiments, the smart driving system 100 and the cloud server communicate wirelessly via a wireless communication network (e.g., a wireless communication network including, but not limited to, a GPRS network, a Zigbee network, a Wifi network, a 3G network, a 4G network, a 5G network, etc.).

In some embodiments, the cloud server is a cloud server established by a vehicle service provider, and provides cloud storage and cloud computing functions. In some embodiments, the cloud server builds the vehicle-side profile. In some embodiments, the vehicle-side profile stores various information uploaded by the intelligent driving system 100. In some embodiments, the cloud server may synchronize the driving data generated by the vehicle side in real time.

In some embodiments, the cloud server may be a server or a server group. The server group may be centralized or distributed. The distributed servers are beneficial to the distribution and optimization of tasks in a plurality of distributed servers, and the defects of resource shortage and response bottleneck of the traditional centralized server are overcome. In some embodiments, the cloud server may be local or remote.

In some embodiments, the cloud server may be used to perform parking charges, road passing charges, etc. for the vehicle end. In some embodiments, the cloud server is further configured to analyze the driving behavior of the driver and perform a safety level assessment on the driving behavior of the driver.

In some embodiments, the cloud server may be used to obtain information about Road monitoring units (RSUs) and smart driving vehicles, and may send the information to the smart driving vehicles. In some embodiments, the cloud server may send detection information corresponding to the smart driving vehicle in the road monitoring unit to the smart driving vehicle according to information of the smart driving vehicle.

In some embodiments, a road monitoring unit may be used to collect road monitoring information. In some embodiments, the road monitoring unit may be an environmental perception sensor, such as a camera, a lidar, etc., and may also be a road device, such as a V2X device, a roadside traffic light device, etc. In some embodiments, the road monitoring units may monitor road conditions pertaining to the respective road monitoring units, e.g., by type of vehicle, speed, priority level, etc. The road monitoring unit can send the road monitoring information to the cloud server after collecting the road monitoring information, and can also send the intelligent driving vehicle through the road.

And the vehicle bottom layer execution system is used for receiving the vehicle control instruction and realizing the control of vehicle running. In some embodiments, vehicle under-floor execution systems include, but are not limited to: a steering system, a braking system and a drive system. The steering system, the braking system and the driving system belong to mature systems in the field of vehicles, and are not described in detail herein.

In some embodiments, the smart-drive vehicle may also include a vehicle CAN bus, not shown in FIG. 1, that connects to the vehicle's underlying implement system. Information interaction between the intelligent driving system 100 and the vehicle bottom layer execution system is transmitted through a vehicle CAN bus.

In some embodiments, the intelligent driving vehicle may control the vehicle to travel by both the driver and the intelligent driving system 100. In the manual driving mode, the driver drives the vehicle by operating devices for controlling the vehicle to run, such as, but not limited to, a brake pedal, a steering wheel, an accelerator pedal, and the like. The device for controlling the vehicle to run can directly operate the vehicle bottom layer execution system to control the vehicle to run.

In some embodiments, the intelligent driving vehicle may also be an unmanned vehicle, and the driving control of the vehicle is performed by the intelligent driving system 100.

Fig. 2 is a block diagram of an intelligent driving system 200 according to an embodiment of the present disclosure. In some embodiments, the intelligent driving system 200 may be implemented as the intelligent driving system 100 of fig. 1 or a part of the intelligent driving system 100 for controlling the vehicle to run.

As shown in fig. 2, the smart driving system 200 may be divided into a plurality of modules, for example, may include: perception module 201, planning module 202, control module 203, map processing module 204, and other modules that may be used for intelligent driving.

The sensing module 201 is used for sensing and positioning the environment. In some embodiments, the sensing module 201 is used to obtain sensor data, V2X (Vehicle to X) data, high precision maps, and the like. In some embodiments, the sensing module 201 is configured to sense and locate the environment based on at least one of acquired sensor data, V2X (Vehicle to X) data, high-precision maps, and the like.

In some embodiments, the sensing module 201 is configured to generate sensing and positioning information, so as to sense an obstacle, identify a travelable area of a camera image, position a vehicle, and the like.

Environmental awareness (Environmental awareness) may be understood as a semantic classification of data with respect to the context of the scene understanding capabilities of the environment, such as the location of obstacles, the detection of road signs/markers, the detection of pedestrians/vehicles, etc. In some embodiments, the environmental sensing may be performed by fusing data of various sensors such as a camera, a laser radar, and a millimeter wave radar.

Localization (Localization) is part of the perception, and is the ability to determine the position of an intelligent driving vehicle relative to the environment. The positioning can be as follows: GPS positioning, wherein the positioning accuracy of the GPS is in the order of tens of meters to centimeters, and the positioning accuracy is high; the positioning method combining the GPS and the Inertial Navigation System (Inertial Navigation System) can also be used for positioning. The positioning may also be performed by using a SLAM (Simultaneous Localization And Mapping), where the target of the SLAM is to construct a map And to perform positioning using the map, And the SLAM determines the position of the current vehicle And the position of the current observed feature by using the environmental features that have been observed.

The V2X is a key technology of the intelligent transportation system, so that the vehicles, the vehicles and the base stations can communicate with each other, a series of traffic information such as real-time road conditions, road information and pedestrian information can be obtained, the intelligent driving safety is improved, the congestion is reduced, the traffic efficiency is improved, and vehicle-mounted entertainment information is provided.

The high accuracy map is the geographical map that uses in the intelligent driving field, compares with traditional map, and the difference lies in: 1) high-precision maps comprise a large amount of driving assistance information, for example by means of an accurate three-dimensional representation of the road network: including intersection places, landmark positions, and the like; 2) high-precision maps also include a large amount of semantic information, such as reporting the meaning of different colors on traffic lights, in turn, for example, indicating the speed limit of roads, and the location where left-turn lanes begin; 3) the high-precision map can reach centimeter-level precision, and the safe driving of the intelligent driving vehicle is ensured.

The planning module 202 is configured to perform path planning and decision making based on the perceptual positioning information generated by the perceptual module 201.

In some embodiments, the planning module 202 is configured to perform path planning and decision-making based on the perceptual positioning information generated by the perception module 201, in combination with at least one of V2X data, high-precision maps, and the like.

In some embodiments, the planning module 202 is used to plan a path, deciding: the planning decision information is generated based on the behavior (e.g., including but not limited to following, passing, parking, detouring, etc.), vehicle heading, vehicle speed, desired acceleration of the vehicle, desired steering wheel angle, etc.

The control module 203 is configured to perform path tracking and trajectory tracking based on the planning decision information generated by the planning module 202.

In some embodiments, the control module 203 is configured to generate control commands for the vehicle floor-based execution system and issue the control commands, so that the vehicle floor-based execution system controls the vehicle to travel according to a desired path, for example, controls the steering wheel, the brake, and the throttle to control the vehicle laterally and longitudinally.

In some embodiments, the control module 203 is further configured to calculate a front wheel steering angle based on a path tracking algorithm.

In some embodiments, the expected path curve in the path tracking process is independent of time parameters, and during tracking control, the intelligent driving vehicle can be assumed to advance at a constant speed at the current speed, so that the driving path approaches to the expected path according to a certain cost rule; during track tracking, the expected path curve is related to both time and space, and the intelligent driving vehicle is required to reach a certain preset reference path point within a specified time.

Path tracking differs from trajectory tracking in that it is not subject to time constraints and only requires the desired path to be tracked within a certain error range.

The map processing module 204 is used for acquiring images and vehicle positioning information acquired by the vision sensor; identifying characters in the image; and determining global coordinates of a plurality of key features corresponding to the characters based on the vehicle positioning information. In some embodiments, the map processing module 204 is used for special scenarios with character semantic uniqueness, such as: parking lots, city roads with signboards, highways, and the like.

In some embodiments, the function of the map processing module 204 may be integrated into the perception module 201, the planning module 202, or the control module 203, or may be configured as a module separate from the intelligent driving system 200, and the map processing module 204 may be a software module, a hardware module, or a module combining software and hardware. For example, the map processing module 204 is a software module running on an operating system, and the in-vehicle hardware system is a hardware system supporting the operating system.

Fig. 3 is a block diagram of an intelligent driving system 300 according to an embodiment of the present disclosure. In some embodiments, the intelligent driving system 300 may be implemented as the intelligent driving system 100 of fig. 1 or a part of the intelligent driving system 100 for controlling the vehicle to run.

As shown in fig. 3, the smart driving system 300 may be divided into a plurality of modules, for example, may include: a perception module 301, a planning module 302, a control module 303, a map processing module 304, a repositioning module 305, and some other modules that may be used for intelligent driving. In some embodiments, the perception module 301, the planning module 302, the control module 303, and the map processing module 304 may be implemented as the perception module 201, the planning module 202, the control module 203, and the map processing module 204, respectively, in fig. 2.

A repositioning module 305, configured to obtain an image acquired by a visual sensor in a special scene; identifying characters in the image; matching the characters with characters inserted in advance in the electronic map; and performing relocation based on the matching result. Wherein, the special scene has character semantic uniqueness, and the special scene is for example: parking lots, city roads with signboards, highways, and the like.

In some embodiments, the functions of the relocation module 305 may be integrated into the perception module 301, the planning module 302, the control module 303, or the map processing module 304, or may be configured as a separate module from the intelligent driving system 300, and the relocation module 305 may be a software module, a hardware module, or a module combining software and hardware. For example, the relocation module 305 is a software module running on an operating system, and the in-vehicle hardware system is a hardware system supporting the running of the operating system.

Fig. 4 is a block diagram of a map processing module 400 provided in an embodiment of the present disclosure. In some embodiments, the map processing module 400 may be implemented as the map processing module 204 or as part of the map processing module 204 in fig. 2.

As shown in fig. 4, the map processing module 400 may include, but is not limited to, the following elements: an acquisition unit 401, a recognition unit 402, a determination unit 403, and an insertion unit 404.

The acquiring unit 401 is configured to acquire an image and vehicle positioning information acquired by the vision sensor. In some embodiments, the obtaining unit 401 obtains the image captured by the vision sensor and the vehicle positioning information after the vehicle runs in a special scene, and the image captured by the vision sensor includes characters because the special scene has character semantic uniqueness, and the character semantic uniqueness has scene uniqueness. The character semantics can be any information with scene uniqueness, including numbers, letters, and words, such as the bin number character "C081" in the ground bin scene shown in fig. 9.

A recognition unit 402 for recognizing characters in the image captured by the vision sensor. In some embodiments, the recognition unit 402 first determines a character region in the image and then recognizes the character in the character region. The character area is understood to be the area in which the character is located. In this embodiment, the recognition characters do not directly recognize the entire image, but the character areas are determined first and then the characters in the character areas are recognized, so that the recognition efficiency is improved.

In some embodiments, the recognition unit 402 determines the character area in the image, specifically: extracting a plurality of key features in the image, wherein the key features can be points, lines, rectangles and other geometric shapes, a relative geometric relationship is preset between the key features and the characters, and the key features are used for positioning character areas; and then extracting the character area based on the plurality of key features and a preset relative geometric relationship. In some embodiments, the manner of extracting the key features may be a conventional manner based on image processing, and may also be a manner based on Machine Learning (ML), where the Machine Learning may be Deep Learning (DL).

In the ground library scene shown in fig. 9 and the urban road scene shown in fig. 10, each circle represents a key feature (key point), and as can be seen, in the ground library scene shown in fig. 9, a corner point of each library location is a key feature; in the urban road scene shown in fig. 10, four corner points of the guideboard are four key features. In fig. 9, for the library location C081, the library location number character "C081" is located at a fixed position of the library location, and therefore, the relative geometric relationship between the library location number character "C081" and two key features of the library location may be predetermined, for example, by measuring in advance. In fig. 10, the relative geometric relationship between the characters "baolute" in the guideboard and the key features of the guideboard may also be predetermined.

In some embodiments, the recognition unit 402 recognizes characters in the character region, and may employ a deep learning based target detection algorithm; deep learning based character recognition algorithms may also be employed. The target detection algorithm realizes character recognition by detecting the position and the category of a single character; the character recognition algorithm may directly output the recognized character string.

In some embodiments, the identifying unit 402 identifies the characters in the character area, specifically: and obtaining characters in the character area based on the character area and a pre-trained machine learning model. In this embodiment, the input of the machine learning model is a character region, and the output is a character in the character region. The machine learning model may be trained using pre-labeled samples, the character regions of which and the characters in the character regions are known information. Compared with the prior art (a feature point method and a direct method) which utilize low-level features of an image and have poor robustness to scene changes such as illumination changes, the embodiment adopts a machine learning method, only a character region needs to be provided, characters in the character region can be obtained through a machine learning model, the influence of the scene changes is avoided, and the robustness to a dynamic environment is strong.

In some embodiments, the recognition unit 402 recognizes characters in the image captured by the visual sensor, specifically: and obtaining characters in the character region and a plurality of key features corresponding to the characters based on the images and a pre-trained machine learning model. In this embodiment, the input of the machine learning model is an image, and the output is a character in the character region and a corresponding key feature. The machine learning model can be trained by using a pre-labeled sample, wherein the character region in the sample, the character in the character region and a plurality of key features corresponding to the character are all known information. It should be noted that, after the machine learning model is trained, the character region is invisible to the outside of the model, for example, to the user, and the user can only know that the character and the corresponding key feature are output by the machine learning model. If the map library shown in fig. 9 is input into the machine learning model, the machine learning model outputs two key features corresponding to the character "C081" and the character "C081" for the library position C081. In the embodiment, the machine learning method is adopted, only the image needs to be provided, the characters in the character region and the corresponding key features can be obtained through the machine learning model, the influence of scene change is avoided, and the robustness to the dynamic environment is strong.

The determining unit 403 is configured to determine global coordinates of a plurality of key features corresponding to the characters based on the vehicle positioning information. In some embodiments, after the determining unit 403 determines that the same character is recognized by the recognizing unit 402 through continuous multi-frame images, the global coordinates of a plurality of key features corresponding to the character are determined based on the vehicle positioning information. The number of consecutive multiframes can be set based on actual needs, and the specific value of the consecutive multiframes is not limited in this embodiment. If the false detection rate of the single-frame image of the identification unit 402 is p, and 0 < p < 1, the false detection rate of the continuous multi-frame image is p^NAnd N is the number of consecutive multiframes. Taking p as 0.1 and N as 3 as an example, the false detection rate of the continuous multi-frame image is 0.001, and it is seen that the false detection rate is reduced by the character recognition of the continuous multi-frame image.

In some embodiments, the determining unit 403 is further configured to establish a mapping index of the character and the plurality of key features after the identifying unit 402 identifies the character in the image. In some embodiments, after the determining unit 403 determines that the same character is recognized by the identifying unit 402 in the continuous multi-frame images, a mapping index of the character and the plurality of key features is established. In some embodiments, the mapping index of the character and the plurality of key features may be implemented using a data structure such as a binary tree or hash. Compared with the prior art (a feature point method and a direct method), the global map needs to be traversed and matched during relocation, and the time efficiency is not high.

In some embodiments, the determining unit 403 is further configured to sort the plurality of key features corresponding to the characters after the identifying unit 402 identifies the characters in the image. In some embodiments, after the determining unit 403 determines that the multiple continuous frames of images of the identifying unit 402 all identify the same character, the multiple key features corresponding to the characters are sorted. In some embodiments, the plurality of key features corresponding to the characters may be sorted based on the geometric relationship inherent in the scene, for example, in the ground library scene shown in fig. 9, the library number character "C081" associates two key features of the library, and the key features closer to "C" may be sorted into 1 and the key features farther from "C" may be sorted into 2. In this embodiment, by sorting the plurality of key features corresponding to the characters, the key features can be matched quickly in the relocation process, and the relocation efficiency is further improved.

In some embodiments, the determining unit 403 determines, based on the vehicle positioning information, global coordinates of a plurality of key features corresponding to the characters, specifically: local coordinates of the plurality of key features in the image are converted to global coordinates based on the vehicle localization information. Wherein, the local coordinates of the key feature in the image can be understood as the pixel coordinates of the key feature in the image. In some embodiments, the local coordinates are converted to global coordinates by:

P_w＝Twv×P_l

where P _ w is the global coordinate of the key feature, P _ l is the local coordinate of the key feature in the image, and Twv is the vehicle location information. For three-dimensional space, Twv is a 4 × 4 transform matrix. The vehicle location information may originate from any location source, such as GPS, SLAM, etc.

And an inserting unit 404, configured to insert global coordinates of the character and the plurality of key features into the electronic map. Wherein an insertion is understood to be a saving into an electronic map.

In some embodiments, the division of each unit in the map processing module 400 is only one logical function division, and there may be another division manner when the actual implementation is performed, for example, the obtaining unit 401, the identifying unit 402, the determining unit 403, and the inserting unit 404 may be implemented as one unit; the obtaining unit 401, the identifying unit 402, the determining unit 403 or the inserting unit 404 may also be divided into a plurality of sub-units. It will be understood that the various units or sub-units may be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application.

Fig. 5 is a block diagram of a relocation module 500 according to an embodiment of the present disclosure. In some embodiments, the relocation module 500 may be implemented as the relocation module 305 in FIG. 3 or as part of the relocation module 305.

As shown in FIG. 5, the relocation module 500 may include, but is not limited to, the following elements: an acquisition unit 501, a recognition unit 502, a matching unit 503 and a relocation unit 504.

An obtaining unit 501 is configured to obtain an image collected by a vision sensor. In some embodiments, the obtaining unit 501 obtains the image collected by the vision sensor after the vehicle is driven in a special scene and the vehicle positioning information cannot be obtained, that is, after it is determined that the vehicle needs to perform special scene repositioning. Because the special scene has the uniqueness of the character semantics, the image collected by the visual sensor comprises the characters, and the character semantics has the uniqueness of the scene. The character semantics can be any information with scene uniqueness, including numbers, letters, and words, such as the bin number character "C081" in the ground bin scene shown in fig. 9.

The recognition unit 502 is used for recognizing characters in the image acquired by the vision sensor. In some embodiments, the recognition unit 502 first determines a character region in the image and then recognizes the character in the character region. The character area is understood to be the area in which the character is located. In this embodiment, the recognition characters do not directly recognize the entire image, but the character areas are determined first and then the characters in the character areas are recognized, so that the recognition efficiency is improved.

In some embodiments, the recognition unit 502 determines the character area in the image, specifically: extracting a plurality of key features in the image, wherein a relative geometric relationship is preset between the key features and the characters, and the key features are used for positioning character areas; and then extracting the character area based on the plurality of key features and a preset relative geometric relationship. In some embodiments, the manner of extracting the key features may be a conventional manner based on image processing, and may also be a manner based on machine learning, where the machine learning may be deep learning.

In some embodiments, the recognition unit 502 recognizes characters in the character region, and may employ a deep learning based target detection algorithm; deep learning based character recognition algorithms may also be employed. The target detection algorithm realizes character recognition by detecting the position and the category of a single character; the character recognition algorithm may directly output the recognized character string.

In some embodiments, the identifying unit 502 identifies characters in the character area, specifically: and obtaining characters in the character area based on the character area and a pre-trained machine learning model. In this embodiment, the input of the machine learning model is a character region, and the output is a character in the character region. The machine learning model may be trained using pre-labeled samples, the character regions of which and the characters in the character regions are known information. Compared with the prior art (a feature point method and a direct method) which utilize low-level features of an image and have poor robustness to scene changes such as illumination changes, the embodiment adopts a machine learning method, only a character region needs to be provided, characters in the character region can be obtained through a machine learning model, the influence of the scene changes is avoided, and the robustness to a dynamic environment is strong.

In some embodiments, the recognition unit 502 recognizes characters in the image captured by the visual sensor, specifically: and obtaining characters in the character region and a plurality of key features corresponding to the characters based on the images and a pre-trained machine learning model. In this embodiment, the input of the machine learning model is an image, and the output is a character in the character region and a corresponding key feature. The machine learning model can be trained by using a pre-labeled sample, wherein the character region in the sample, the character in the character region and a plurality of key features corresponding to the character are all known information. It should be noted that, after the machine learning model is trained, the character region is invisible to the outside of the model, for example, to the user, and the user can only know that the character and the corresponding key feature are output by the machine learning model. If the map library shown in fig. 9 is input into the machine learning model, the machine learning model outputs two key features corresponding to the character "C081" and the character "C081" for the library position C081. In the embodiment, the machine learning method is adopted, only the image needs to be provided, the characters in the character region and the corresponding key features can be obtained through the machine learning model, the influence of scene change is avoided, and the robustness to the dynamic environment is strong.

A matching unit 503, configured to match the characters in the image with the characters inserted in the electronic map in advance. In some embodiments, after the matching unit 503 determines that the same character is recognized by the recognition unit 502 in the continuous multi-frame images, the character in the images is matched with the character inserted in advance in the electronic map. The number of consecutive multiframes can be set based on actual needs, and the specific value of the consecutive multiframes is not limited in this embodiment. If the false detection rate of a single frame image of the identification unit 502 is p, and 0 < p < 1, the false detection rate of a continuous multi-frame image is p^NAnd N is the number of consecutive multiframes. Taking p as 0.1 and N as 3 as an example, the false detection rate of the continuous multi-frame image is 0.001, and it is seen that the false detection rate is reduced by the character recognition of the continuous multi-frame image.

In some embodiments, in the process of processing the electronic map, a data structure such as a binary tree or a hash is used to implement mapping indexes of characters and a plurality of key features, so that compared with the prior art (a feature point method and a direct method), traversal matching needs to be performed on the global map during relocation, and time efficiency is not high; if the matching fails, the relocation fails; if the matching is successful, relocation can be performed.

A relocation unit 504 for performing relocation based on the matching result. In some embodiments, after the relocation unit 504 determines a match, the relocation unit may index global coordinates of a plurality of key features corresponding to characters in the image based on global coordinates of key features previously inserted in the electronic map; and then repositioning is performed based on the global coordinates of the plurality of key features.

In some embodiments, the repositioning unit 504 indexes global coordinates of a plurality of key features corresponding to characters in the image, specifically: indexing a plurality of key features matched with a plurality of key features corresponding to characters in an image based on a mapping index of a pre-established character and the plurality of key features; and further determining the global coordinates of a plurality of key features corresponding to the characters in the image based on the global coordinates of the key features inserted in the electronic map in advance.

In some embodiments, since the key features are sorted during the processing of the electronic map, the repositioning unit 504 determines global coordinates of a plurality of key features corresponding to characters in the image based on global coordinates of key features inserted in the electronic map in advance, specifically: sorting a plurality of key features corresponding to characters in the image; and further determining the global coordinates of a plurality of key features corresponding to the characters in the image based on the sorting result and the global coordinates of the key features inserted in the electronic map in advance. For example, in the ground library scenario shown in fig. 9, the library position number character "C081" is associated with two key features of the library position, and the key feature closer to "C" may be ranked as 1, and the key feature farther from "C" may be ranked as 2. The repositioning unit 504 determines the global coordinates of the key feature ordered as 1 of the two key features corresponding to "C081" pre-inserted in the electronic map as the global coordinates of the key feature ordered as 1 of the two key features corresponding to "C081" in the image.

In some embodiments, the relocation unit 504 performs relocation based on global coordinates of a plurality of key features, specifically: vehicle localization information is determined based on global coordinates of the plurality of key features and local coordinates of the plurality of key features in the image. Wherein, the local coordinates of the key feature in the image can be understood as the pixel coordinates of the key feature in the image. In some embodiments, the repositioning unit 504 determines the vehicle positioning information based on the global coordinates of the plurality of key features and the local coordinates of the plurality of key features in the image, using a filter-based method, such as a particle filtering method, or using a method based on non-linear optimization, etc.

In some embodiments, the repositioning unit 504 will determine the vehicle location information by:

Twv*＝arg min∑||Twv×P_l－P_w||

in the embodiment, P _ w is a global coordinate of the key feature, P _ l is a local coordinate of the key feature in the image, Twv is vehicle positioning information, Twv is vehicle positioning information to be solved, and Twv is an optimal solution of Twv. For three-dimensional space, Twv is a 4 × 4 transform matrix. In this embodiment, multiple sets of P _ l and P _ w are known, and the optimal solution Twv of Twv is obtained, generally using the least square method.

The relocation module 500 shown in fig. 5 does not need to manually post an enhancement pattern, and thus, does not increase labor cost or damage the original scene.

In some embodiments, the division of each unit in the relocation module 500 is only one logical function division, and there may be another division manner in actual implementation, for example, the obtaining unit 501, the identifying unit 502, the matching unit 503, and the relocation unit 504 may be implemented as one unit; the acquisition unit 501, the identification unit 502, the matching unit 503 or the relocation unit 504 may also be divided into a plurality of sub-units. It will be understood that the various units or sub-units may be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application.

Fig. 6 is a schematic structural diagram of an in-vehicle device provided in an embodiment of the present disclosure. The vehicle-mounted equipment can support the operation of the intelligent driving system.

As shown in fig. 6, the vehicle-mounted apparatus includes: at least one processor 601, at least one memory 602, and at least one communication interface 603. The various components in the in-vehicle device are coupled together by a bus system 604. A communication interface 603 for information transmission with an external device. Understandably, the bus system 604 is used to enable connective communication between these components. The bus system 604 includes a power bus, a control bus, and a status signal bus in addition to a data bus. But for the sake of clarity the various busses are labeled in fig. 6 as the bus system 604.

It will be appreciated that the memory 602 in this embodiment can be either volatile memory or nonvolatile memory, or can include both volatile and nonvolatile memory.

In some embodiments, memory 602 stores the following elements, executable units or data structures, or a subset thereof, or an expanded set thereof: an operating system and an application program.

The operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, and the like, and is used for implementing various basic services and processing hardware-based tasks. The application programs, including various application programs such as a Media Player (Media Player), a Browser (Browser), etc., are used to implement various application services. A program for implementing the method for processing an electronic map or the method for visually repositioning a vehicle provided by the embodiments of the present disclosure may be included in an application program.

In the embodiment of the present disclosure, the processor 601 is configured to execute the steps of the method for processing an electronic map or the method for visually repositioning a vehicle provided by the embodiment of the present disclosure by calling a program or an instruction stored in the memory 602, which may be, in particular, a program or an instruction stored in an application program.

The method for processing the electronic map or the method for visually repositioning the vehicle provided by the embodiment of the disclosure can be applied to the processor 601 or implemented by the processor 601. The processor 601 may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware or instructions in the form of software in the processor 601. The Processor 601 may be a general-purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic device, or discrete hardware components. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The steps of the method for processing the electronic map or the method for visually repositioning the vehicle provided by the embodiment of the disclosure can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software units in the decoding processor. The software elements may be located in ram, flash, rom, prom, or eprom, registers, among other storage media that are well known in the art. The storage medium is located in a memory 602, and the processor 601 reads the information in the memory 602 and performs the steps of the method in combination with its hardware.

Fig. 7 is a flowchart of a processing method of an electronic map according to an embodiment of the present disclosure. The execution subject of the method is the vehicle-mounted equipment, and in some embodiments, the execution subject of the method is an intelligent driving system supported by the vehicle-mounted equipment.

As shown in fig. 7, the processing method of the electronic map may include the following steps 701 to 704:

701. and acquiring images and vehicle positioning information acquired by the vision sensor. In some embodiments, the image captured by the vision sensor and the vehicle positioning information are acquired after the vehicle is driven in a special scene, and the image captured by the vision sensor comprises characters due to the fact that the special scene has character semantic uniqueness, and the character semantic uniqueness has scene uniqueness. The character semantics can be any information with scene uniqueness, including numbers, letters, and words, such as the bin number character "C081" in the ground bin scene shown in fig. 9.

702. Characters in the image are identified. In some embodiments, a character region in the image is first determined, and then characters in the character region are identified. The character area is understood to be the area in which the character is located. In this embodiment, the recognition characters do not directly recognize the entire image, but the character areas are determined first and then the characters in the character areas are recognized, so that the recognition efficiency is improved.

In some embodiments, the character region in the image is determined, specifically: extracting a plurality of key features in the image, wherein a relative geometric relationship is preset between the key features and the characters, and the key features are used for positioning character areas; and then extracting the character area based on the plurality of key features and a preset relative geometric relationship. In some embodiments, the manner of extracting the key features may be a conventional manner based on image processing, and may also be a manner based on machine learning, where the machine learning may be deep learning.

In some embodiments, characters in the character region are identified, and a deep learning based target detection algorithm may be employed; deep learning based character recognition algorithms may also be employed. The target detection algorithm realizes character recognition by detecting the position and the category of a single character; the character recognition algorithm may directly output the recognized character string.

In some embodiments, the characters in the character region are identified by: and obtaining characters in the character area based on the character area and a pre-trained machine learning model. In this embodiment, the input of the machine learning model is a character region, and the output is a character in the character region. The machine learning model may be trained using pre-labeled samples, the character regions of which and the characters in the character regions are known information. Compared with the prior art (a feature point method and a direct method) which utilize low-level features of an image and have poor robustness to scene changes such as illumination changes, the embodiment adopts a machine learning method, only a character region needs to be provided, characters in the character region can be obtained through a machine learning model, the influence of the scene changes is avoided, and the robustness to a dynamic environment is strong.

In some embodiments, identifying characters in an image captured by a vision sensor specifically includes: and obtaining characters in the character region and a plurality of key features corresponding to the characters based on the images and a pre-trained machine learning model. In this embodiment, the input of the machine learning model is an image, and the output is a character in the character region and a corresponding key feature. The machine learning model can be trained by using a pre-labeled sample, wherein the character region in the sample, the character in the character region and a plurality of key features corresponding to the character are all known information. It should be noted that, after the machine learning model is trained, the character region is invisible to the outside of the model, for example, to the user, and the user can only know that the character and the corresponding key feature are output by the machine learning model. If the map library shown in fig. 9 is input into the machine learning model, the machine learning model outputs two key features corresponding to the character "C081" and the character "C081" for the library position C081. In the embodiment, the machine learning method is adopted, only the image needs to be provided, the characters in the character region and the corresponding key features can be obtained through the machine learning model, the influence of scene change is avoided, and the robustness to the dynamic environment is strong.

703. And determining global coordinates of a plurality of key features corresponding to the characters based on the vehicle positioning information. In some embodiments, after determining that the continuous multi-frame images all identify the same character, determining global coordinates of a plurality of key features corresponding to the character based on the vehicle positioning information. The number of consecutive multiframes can be set based on actual needs, and the specific value of the consecutive multiframes is not limited in this embodiment. If the false detection rate of the single-frame image is p, and p is more than 0 and less than 1, the false detection rate of the continuous multi-frame image is p^NAnd N is the number of consecutive multiframes. Taking p as 0.1 and N as 3 as an example, the false detection rate of the continuous multi-frame image is 0.001, and it is seen that the false detection rate is reduced by the character recognition of the continuous multi-frame image.

In some embodiments, after identifying the character in the image, a mapping index of the character to a plurality of key features is established. In some embodiments, after determining that the continuous multi-frame images all identify the same character, establishing a mapping index of the character and a plurality of key features. In some embodiments, the mapping index of the character and the plurality of key features may be implemented using a data structure such as a binary tree or hash. Compared with the prior art (a feature point method and a direct method), the global map needs to be traversed and matched during relocation, and the time efficiency is not high.

In some embodiments, after identifying the characters in the image, a plurality of key features corresponding to the characters are sorted. In some embodiments, after determining that the continuous multi-frame images all identify the same character, the plurality of key features corresponding to the character are sorted. In some embodiments, the plurality of key features corresponding to the characters may be sorted based on the geometric relationship inherent in the scene, for example, in the ground library scene shown in fig. 9, the library number character "C081" associates two key features of the library, and the key features closer to "C" may be sorted into 1 and the key features farther from "C" may be sorted into 2. In this embodiment, by sorting the plurality of key features corresponding to the characters, the key features can be matched quickly in the relocation process, and the relocation efficiency is further improved.

In some embodiments, based on the vehicle positioning information, global coordinates of a plurality of key features corresponding to the characters are determined, specifically: local coordinates of the plurality of key features in the image are converted to global coordinates based on the vehicle localization information. Wherein, the local coordinates of the key feature in the image can be understood as the pixel coordinates of the key feature in the image. In some embodiments, the local coordinates are converted to global coordinates by:

P_w＝Twv×P_l

704. Inserting the global coordinates of the character and the plurality of key features into an electronic map. Wherein an insertion is understood to be a saving into an electronic map.

FIG. 8 is a flowchart of a method for visual repositioning of a vehicle according to an embodiment of the present disclosure. The execution subject of the method is the vehicle-mounted equipment, and in some embodiments, the execution subject of the method is an intelligent driving system supported by the vehicle-mounted equipment.

As shown in fig. 8, a method of vehicle vision repositioning may include the following steps 801 to 804:

801. and acquiring an image acquired by the vision sensor. In some embodiments, the images captured by the vision sensor are acquired after the vehicle is traveling in a particular scene and the vehicle positioning information cannot be acquired, i.e., after it is determined that the vehicle needs to be repositioned in the particular scene. Because the special scene has the uniqueness of the character semantics, the image collected by the visual sensor comprises the characters, and the character semantics has the uniqueness of the scene. The character semantics can be any information with scene uniqueness, including numbers, letters, and words, such as the bin number character "C081" in the ground bin scene shown in fig. 9.

802. Characters in the image are identified. In some embodiments, a character region in the image is first determined, and then characters in the character region are identified. The character area is understood to be the area in which the character is located. In this embodiment, the recognition characters do not directly recognize the entire image, but the character areas are determined first and then the characters in the character areas are recognized, so that the recognition efficiency is improved.

803. And matching the characters in the image with the characters inserted in the electronic map in advance. In some embodiments, after determining that the continuous multi-frame images all identify the same character, matching the character in the images with the character pre-inserted in the electronic map. The number of consecutive multiframes can be set based on actual needs, and the specific value of the consecutive multiframes is not limited in this embodiment. If the false detection rate of the single-frame image is p, and p is more than 0 and less than 1, the false detection rate of the continuous multi-frame image is p^NAnd N is the number of consecutive multiframes. Taking p as 0.1 and N as 3 as an example, the false detection rate of the continuous multi-frame image is 0.001, and it is seen that the false detection rate is reduced by the character recognition of the continuous multi-frame image.

In some embodiments, because in the process of processing the electronic map, a data structure such as a binary tree or a hash is adopted to realize mapping indexes of characters and a plurality of key features, compared with the prior art (a feature point method and a direct method), the time efficiency is not high when the global map needs to be traversed and matched during relocation, in this embodiment, characters matched with characters in an image can be inquired in the electronic map more quickly; if the matching fails, the relocation fails; if the matching is successful, relocation can be performed.

804. And performing relocation based on the matching result. In some embodiments, after determining the match, the global coordinates of a plurality of key features corresponding to the characters in the image may be indexed based on global coordinates of key features previously inserted in the electronic map; and then repositioning is performed based on the global coordinates of the plurality of key features.

In some embodiments, the global coordinates of the plurality of key features corresponding to the characters in the index image are specifically: indexing a plurality of key features matched with a plurality of key features corresponding to characters in an image based on a mapping index of a pre-established character and the plurality of key features; and further determining the global coordinates of a plurality of key features corresponding to the characters in the image based on the global coordinates of the key features inserted in the electronic map in advance.

In some embodiments, since the key features are sorted in the processing process of the electronic map, the global coordinates of a plurality of key features corresponding to characters in the image are determined based on the global coordinates of the key features inserted in the electronic map in advance, specifically: sorting a plurality of key features corresponding to characters in the image; and further determining the global coordinates of a plurality of key features corresponding to the characters in the image based on the sorting result and the global coordinates of the key features inserted in the electronic map in advance. For example, in the ground library scenario shown in fig. 9, the library position number character "C081" is associated with two key features of the library position, and the key feature closer to "C" may be ranked as 1, and the key feature farther from "C" may be ranked as 2. Therefore, the global coordinates of the key feature ranked as 1 in the two key features corresponding to "C081" inserted in advance in the electronic map are determined as the global coordinates of the key feature ranked as 1 in the two key features corresponding to "C081" in the image.

In some embodiments, the repositioning is performed based on global coordinates of a plurality of key features, specifically: vehicle localization information is determined based on global coordinates of the plurality of key features and local coordinates of the plurality of key features in the image. Wherein, the local coordinates of the key feature in the image can be understood as the pixel coordinates of the key feature in the image. In some embodiments, the vehicle positioning information may be determined using a filter-based method, such as a particle filtering method, a non-linear optimization-based method, and the like, based on the global coordinates of the plurality of key features and the local coordinates of the plurality of key features in the image.

In some embodiments, vehicle location information will be determined by:

Twv*＝arg min∑||Twv×P_l－P_w||

The method for visually repositioning the vehicle shown in fig. 8 does not need to manually paste the enhancement patterns, does not increase the labor cost, and does not damage the original scene.

It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of action combinations, but those skilled in the art will understand that the present disclosure embodiments are not limited by the described action sequences, because some steps may be performed in other sequences or simultaneously according to the present disclosure embodiments (for example, during the process of processing an electronic map, the image acquired by the vision sensor and the vehicle positioning information may be performed simultaneously, or the vehicle positioning information may be acquired after the mapping index of the characters and the plurality of key features is established). In addition, those skilled in the art can appreciate that the embodiments described in the specification all belong to alternative embodiments.

Embodiments of the present disclosure also provide a non-transitory computer-readable storage medium storing a program or instructions, where the program or instructions cause a computer to perform steps of various embodiments of a processing method such as an electronic map or a vehicle vision repositioning method, and are not repeated herein to avoid repeated descriptions.

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

Those skilled in the art will appreciate that while some embodiments described herein include some features included in other embodiments, rather than others, combinations of features of different embodiments are meant to be within the scope of the disclosure and form different embodiments.

Those skilled in the art will appreciate that the description of each embodiment has a respective emphasis, and reference may be made to the related description of other embodiments for those parts of an embodiment that are not described in detail.

Although the embodiments of the present disclosure have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present disclosure, and such modifications and variations fall within the scope defined by the appended claims.

Claims

1. A processing method of an electronic map is characterized by comprising the following steps:

identifying characters in the image;

2. The method of claim 1, wherein identifying characters in the image comprises:

determining a character area in the image;

identifying a character in the character area.

3. The method of claim 2, wherein determining a character region in the image comprises:

extracting a plurality of key features in the image; the key features and the characters are preset with relative geometric relations;

and extracting a character region based on the plurality of key features and the relative geometrical relationship.

4. The method of claim 2, wherein identifying the character in the character region comprises:

and obtaining characters in the character area based on the character area and a pre-trained machine learning model.

5. The method of claim 1, wherein identifying characters in the image comprises:

and obtaining characters in the character region and a plurality of key features corresponding to the characters based on the images and a pre-trained machine learning model.

6. The method of any of claims 1 to 5, wherein after identifying the character in the image, the method further comprises:

and after determining that the same characters are identified in the continuous multi-frame images, executing a step of determining the global coordinates of a plurality of key features corresponding to the characters.

7. The method of claim 1, wherein upon identifying a character in the image, the method further comprises:

and establishing a mapping index of the character and a plurality of key features.

8. A method of vehicle vision repositioning for a special scene having character semantic uniqueness, the method comprising:

acquiring an image acquired by a visual sensor;

identifying characters in the image;

matching characters in the image with characters inserted in advance in an electronic map;

and performing relocation based on the matching result.

9. An in-vehicle apparatus, characterized by comprising: a processor and a memory;

the processor is adapted to perform the steps of the method of any one of claims 1 to 8 by calling a program or instructions stored in the memory.

10. A non-transitory computer-readable storage medium storing a program or instructions for causing a computer to perform the steps of the method according to any one of claims 1 to 8.