Disclosure of Invention
Object of the application
Based on this, in order to enable a user to independently and conveniently realize the autonomous transfer between the electric wheelchair and a carrier such as a sofa or a bed without the assistance of other people and simultaneously realize the autonomous movement of the electric wheelchair between a standby space and a target carrier, the following technical scheme is disclosed in the present application.
(II) technical scheme
As a first aspect of the present application, the present application discloses a method for autonomous operation based on an electric wheelchair, comprising:
the electric wheelchair moves to the vicinity of a target carrier by using an autonomous navigation function and automatically determines a reference object;
acquiring three-dimensional point cloud data and an image of a reference object, and processing the three-dimensional point cloud data and the image to obtain spatial position and attitude information of the reference object;
and calculating the position adjustment amount and/or the orientation adjustment amount of the electric wheelchair body according to the spatial position posture information of the reference object, and adjusting the position and/or the orientation of the electric wheelchair body according to the calculation result until the position and/or the orientation meet the requirement that a user transfers between the electric wheelchair and a target carrier.
In one possible embodiment, moving to the vicinity of the object carrier with the autonomous navigation function comprises:
after receiving the moving instruction, the electric wheelchair plans a moving path according to a pre-established three-dimensional indoor map;
the electric wheelchair moves to the vicinity of the target carrier according to the planned moving path;
the electric wheelchair obtains an environment point cloud of the surrounding environment by using a point cloud data acquisition device in the moving process, detects an obstacle based on the environment point cloud, replans a moving path according to the position of the obstacle and moves according to the newly planned moving path.
In one possible embodiment, the step of creating a three-dimensional indoor map comprises:
acquiring indoor three-dimensional point cloud data through point cloud data acquisition equipment of the electric wheelchair;
filtering, registering and dividing the three-dimensional point cloud data to obtain a three-dimensional point cloud map;
and converting the obtained three-dimensional point cloud map into a grid type three-dimensional indoor map.
In one possible implementation, the processing the three-dimensional point cloud data and the image to obtain the spatial position and posture information of the reference object includes:
point cloud data acquisition equipment and a visual sensor are subjected to combined calibration to calculate coordinate conversion parameters;
processing the obtained three-dimensional point cloud data, modeling, and screening out the three-dimensional point cloud data of the reference object to be detected;
calculating an area corresponding to the three-dimensional point cloud data of the reference object to be detected in the image by using the coordinate conversion parameters;
and performing decision fusion on the obtained three-dimensional point cloud data and the image, and comprehensively judging the spatial position and attitude information of the reference object to be detected.
In one possible embodiment, when the user is transferred from the destination carrier to the electric wheelchair, the destination carrier and the user himself are taken as references;
when the user transfers the target carrier from the electric wheelchair, the target carrier is used as a reference object.
As a second aspect of the present application, the present application discloses a system for autonomous operation based on an electric wheelchair, comprising:
the autonomous navigation module is used for realizing autonomous movement of the electric wheelchair to the vicinity of a target carrier and automatically determining a reference object;
the point cloud data acquisition equipment is used for acquiring three-dimensional point cloud data and images of a reference object;
the reference object detection module is used for processing the three-dimensional point cloud data and the image to obtain spatial position and attitude information of a reference object;
the calculation and adjustment module is used for calculating the position adjustment amount and/or the orientation adjustment amount of the electric wheelchair body according to the spatial position posture information of the reference object, and adjusting the position and/or the orientation of the electric wheelchair body according to the calculation result until the position and/or the orientation meet the requirement that a user transfers between the electric wheelchair and the target carrier.
In one possible embodiment, the autonomous navigation module comprises:
the path planning unit is used for planning a moving path according to a pre-established three-dimensional indoor map after receiving the moving instruction;
the control unit is used for sending corresponding driving signals to the driving unit according to the planned moving path;
the driving unit is used for controlling the motor to operate according to the driving signal output by the control unit and driving the electric wheelchair body to move to the position near the target carrier through the motor;
the electric wheelchair obtains an environment point cloud of the surrounding environment by using a point cloud data acquisition device in the moving process, detects an obstacle based on the environment point cloud, replans a moving path according to the position of the obstacle and moves according to the newly planned moving path.
In one possible embodiment, the system further comprises a map creation module, the map creation module comprising:
the data acquisition unit is used for acquiring indoor three-dimensional point cloud data through point cloud data acquisition equipment of the electric wheelchair;
the data processing unit is used for filtering, registering and dividing the three-dimensional point cloud data collected by the data collecting unit to obtain a three-dimensional point cloud map;
the map conversion unit is used for converting the obtained three-dimensional point cloud map into a grid type three-dimensional indoor map.
In one possible embodiment, the system further comprises a map creation module, the map creation module comprising:
the joint calibration unit is used for carrying out joint calibration on the point cloud data acquisition equipment and the visual sensor to calculate coordinate conversion parameters;
the data screening unit is used for processing the acquired three-dimensional point cloud data and then modeling to screen out the three-dimensional point cloud data of the reference object to be detected;
the information conversion unit is used for calculating an area corresponding to the three-dimensional point cloud data of the reference object to be detected in the image by using the coordinate conversion parameters;
and the comprehensive judgment unit is used for performing decision fusion on the obtained three-dimensional point cloud data and the image and comprehensively judging the spatial position and posture information of the reference object to be detected.
In one possible embodiment, when the user is transferred from the destination carrier to the electric wheelchair, the destination carrier and the user himself are taken as references;
when the user transfers the target carrier from the electric wheelchair, the target carrier is used as a reference object.
(III) advantageous effects
According to the method and the device for autonomous operation of the electric wheelchair, the relative position between the reference object and the wheelchair is judged by detecting the position and the posture of the wheelchair, so that the position and the posture of the wheelchair required by autonomous transfer between the reference object and the wheelchair can be calculated by a user, the position and the posture of the wheelchair can be adjusted according to the position and the posture, the user can independently and conveniently realize autonomous transfer between the electric wheelchair and carriers such as a sofa and a bed without assistance of other people, and meanwhile, the electric wheelchair can also realize autonomous movement between a standby space and a target carrier, so that the electric wheelchair is more humanized and intelligent, and the physiological and psychological comfort level of the user is improved.
Detailed Description
In order to make the implementation objects, technical solutions and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be described in more detail below with reference to the drawings in the embodiments of the present application.
Referring to fig. 1, a method for autonomous operation based on an electric wheelchair disclosed in the present application is described in detail, and as shown in fig. 1, the method disclosed in the present embodiment mainly includes the following steps 100 to 300.
And step 100, the electric wheelchair moves to the vicinity of the target carrier by using the autonomous navigation function, and automatically determines a reference object.
The object carrier refers to a carrier in a home space, such as a sofa, a bed, a chair, etc., which can carry the body of a user. If the user wants to realize the autonomous transfer between the electric wheelchair and the target carrier, the electric wheelchair should move to the vicinity of the target carrier first, and then the user performs the transfer action.
The reference object is a reference standard set for a user to calculate the position and orientation of the wheelchair to be adjusted during the autonomous transfer process between the electric wheelchair and the target carrier, and comprises the target carrier and the user.
When a user wants to transfer the target carrier from the electric wheelchair, the user firstly takes the electric wheelchair to automatically move to the vicinity of the target carrier, and the electric wheelchair automatically determines the target carrier which the user wants to transfer as a reference object.
If the user is located in the target carrier and wants to transfer to the electric wheelchair, the electric wheelchair automatically runs to the position near the target carrier, and the user and the target carrier are automatically determined to be reference objects.
And 200, acquiring three-dimensional point cloud data and an image of the reference object, and processing the three-dimensional point cloud data and the image to obtain spatial position and attitude information of the reference object.
After the electric wheelchair automatically determines the reference object, in order to obtain comprehensive and accurate spatial position and attitude information of the reference object, a point cloud data acquisition device and a visual sensor are adopted for joint detection, and firstly, two devices respectively obtain the region information containing the reference object. The point cloud data acquisition equipment acquires three-dimensional point cloud data information of an area containing a reference object, wherein the point cloud data is a set of vectors in a three-dimensional coordinate system. The scan data is recorded in the form of dots, each dot containing three-dimensional coordinates, some of which may contain color information or reflection intensity information. The vision sensor acquires image information of a region including a reference object. In order to better fuse the information acquired by the point cloud data acquisition equipment and the visual sensor, the point cloud data acquisition equipment and the visual sensor are firstly subjected to combined calibration and then the acquired information is processed, and finally the spatial position and attitude information of the reference object is finally determined by integrating two detection results. After the spatial position posture information of the reference object is determined, the electric wheelchair can calculate the position and/or orientation of the electric wheelchair to be adjusted according to the spatial position posture information of the reference object, and then the position and/or orientation of the electric wheelchair to be adjusted according to the calculation result, so that the user can realize the autonomous transfer between the electric wheelchair and the target carrier. The spatial position and orientation information of the reference object includes coordinates, shape, height, angle, orientation, and the like.
And 300, calculating the position adjustment amount and/or the orientation adjustment amount of the electric wheelchair body according to the spatial position posture information of the reference object, and adjusting the position and/or the orientation of the electric wheelchair body according to the calculation result until the position and/or the orientation meet the requirement that the user transfers between the electric wheelchair and the target carrier.
After the spatial position and posture information of the reference object is determined, the electric wheelchair needs to be adjusted in order to realize the autonomous transfer of the user between the electric wheelchair and the target carrier, and calculation needs to be carried out before the adjustment. Firstly, calculating the orientation and the adjustment angle to be adjusted, and firstly, determining the regional center point of the electric wheelchair according to the regional three-dimensional point cloud data information and the regional image information which contain the reference object and are obtained in the step 200; determining the center point of the reference object according to the three-dimensional point cloud data information and the image information of the reference object finally determined in the step 200, and calculating the direction and the angle of the electric wheelchair to be adjusted when the area center point of the reference object is coincident with the reference object center point, namely when the reference object is located in the center of the area; after the orientation and the adjusting angle of the electric wheelchair to be adjusted are determined, the distance to be adjusted is calculated, the adjusting distance is calculated according to the adjusting coefficient, and the optimal adjusting coefficient is the area ratio of the reference object to the area where the reference object is located when the user can realize the autonomous transfer between the electric wheelchair and the target carrier. And calculating the distance which needs to be adjusted by the electric wheelchair when the area ratio of the reference object to the area of the reference object reaches the optimal adjustment coefficient, and adjusting the orientation, the angle and the distance of the electric wheelchair according to the calculation result. After the primary calculation and adjustment are finished, the electric wheelchair judges whether the expected effect of the autonomous transfer of the user between the electric wheelchair and the target carrier can be realized, and if the expected effect can be realized, the adjustment is finished; if not, the calculation and adjustment actions are repeated until the user is able to achieve the desired effect of autonomous transfer between the powered wheelchair and the target carrier, and the adjustment is complete.
In at least one embodiment, the calculation and adjustment actions of the electric wheelchair may be performed once or may be performed a plurality of times in a cycle, that is: and when the expected effect is not achieved after the calculation, adjustment and judgment steps are carried out once, the calculation, adjustment and judgment steps are continued until the judgment result of the electric wheelchair reaches the expected effect, namely, the user can realize the autonomous transfer between the electric wheelchair and the target carrier. Step 300 ends.
Illustratively, when a user wants to effect a transfer from an electric wheelchair to a sofa, the user travels in the vicinity of the sofa in the electric wheelchair. The method comprises the steps of obtaining regional three-dimensional point cloud data information including sofas through point cloud data acquisition equipment, obtaining regional image information including the sofas through a visual sensor, integrating the regional three-dimensional point cloud data information and the regional image information, jointly determining an integral regional image including the sofas, further determining a geometric center point of the integral regional image, processing the integral regional image, screening out sofa regions, further determining the geometric center point of the sofa regions, and calculating the orientation and the orientation rotation angle of the orientation required to be adjusted by the electric wheelchair when the geometric center point of the integral regional image is overlapped with the geometric center point of the sofa regions by the electric wheelchair. And then, calculating the distance, and calculating the adjusting distance according to the adjusting coefficient, wherein the optimal adjusting coefficient can achieve that the electric wheelchair body and the sofa edge are seamless, the height of the wheelchair seat and the height of the sofa are positioned on the same plane, and a user can move to the sofa through translation.
When the user accomplished the transfer action from carriers such as electronic wheelchair to sofa, bed, electronic wheelchair can independently plan the route, returns the standby space to prevent placing at will and occupy the indoor space or shelter from the sight, and can independently switch on in the standby space when the electric quantity is not enough, can also realize the electric quantity and be full of the back and independently cut off the power supply.
In at least one embodiment, the step 100 of utilizing the autonomous navigation function to move to the vicinity of the object carrier further comprises steps 110 to 120.
And step 110, after the electric wheelchair receives the transfer instruction, planning a moving path according to a pre-established three-dimensional indoor map.
When the electric wheelchair is located indoors where a three-dimensional indoor map is created in advance, a user can send an instruction for moving to a target carrier to the electric wheelchair through voice calling or Bluetooth and the like, after the electric wheelchair receives the instruction, modeling is firstly carried out on the surrounding environment of the three-dimensional indoor map created in advance, all feasible routes are planned according to the three-dimensional indoor map, and a shortest route which can be used for allowing the electric wheelchair to freely pass is determined based on a D-x algorithm. The electric wheelchair is located in a room where a three-dimensional indoor map is created in advance, and the electric wheelchair is located in the room where the three-dimensional indoor map is created in advance when a user takes the electric wheelchair and is located in the room where the three-dimensional indoor map is created in advance when the electric wheelchair is in an empty state.
And step 120, the electric wheelchair moves to the vicinity of the target carrier according to the planned moving path.
After a movement path is planned by the electric wheelchair, the electric wheelchair controller sends a driving signal to the driving device according to the planned path, the driving device drives the motor to drive the electric wheelchair body to move according to the set movement path, after the electric wheelchair starts to move, the point cloud data acquisition equipment is started to acquire environmental point cloud of the surrounding environment, obstacle detection is carried out on the basis of the environmental point cloud, and if no obstacle exists, the electric wheelchair moves to the vicinity of a target carrier according to the planned path; and if the obstacle is encountered, replanning the moving path according to the position of the obstacle and moving according to the newly planned moving path until the object carrier is reached.
In at least one embodiment, the electric wheelchair needs to be used for the first time before a map is not created in a room, and a three-dimensional indoor map is created, wherein the creation process of the three-dimensional indoor map comprises steps 111 to 113.
And step 111, acquiring indoor three-dimensional point cloud data through point cloud data acquisition equipment of the electric wheelchair.
The electric wheelchair moves from any position in an indoor environment of unknown environment, a 360-degree panoramic camera of a vision sensor is started, and the current position, the passing area and the non-passing area are identified at any time. The vision sensor can help the electric wheelchair to effectively grasp the surrounding environment. Meanwhile, the electric wheelchair point cloud data acquisition equipment acquires the environmental point cloud of the surrounding environment, and when the point cloud data is collected, obstacle detection is carried out on the basis of the environmental point cloud to avoid obstacles, so that indoor comprehensive movement in an unknown environment and collection of environmental three-dimensional point cloud data are smoothly completed.
The point cloud acquisition equipment can be built by combining two-dimensional point cloud data acquisition equipment and a rotating platform, and can also be directly three-dimensional point cloud acquisition equipment.
And 112, filtering, registering and dividing the three-dimensional point cloud data to obtain a three-dimensional point cloud map.
Firstly, filtering all three-dimensional point cloud data acquired in the step 111, specifically filtering sparse outliers in the three-dimensional point cloud data, secondly, registering the point cloud data, and connecting effective point clouds of different positions subjected to filtering processing into a complete three-dimensional point cloud map by taking a geometric structure as a reference; the complete three-dimensional point cloud map comprises all indoor information such as articles, walls, passages and the like, the walls, the ground and the ceiling are separated firstly by utilizing point cloud segmentation, and then the indoor articles are separated one by one. And obtaining a complete and clear three-dimensional indoor point cloud map.
In at least one embodiment, an effective area is set, distant outliers outside the area are deleted, the outliers in the area are deleted based on statistical analysis, point cloud registration is carried out by using an ICP (inductively coupled plasma) algorithm after filtering is finished, a complete three-dimensional point cloud map is obtained, separation of a wall, a ground and a ceiling is carried out by using a RANSAC (random SAmple consensus) algorithm after the complete three-dimensional point cloud map is obtained, a better visual effect is obtained, and indoor articles are separated by European clustering.
And 113, converting the obtained three-dimensional point cloud map into a grid type three-dimensional indoor map.
The obtained three-dimensional point cloud map is converted into a three-dimensional grid map by using the OctMap, the size of the grid can be adjusted according to the size of the indoor space, and the obtained three-dimensional grid map can be used for planning a moving path more easily.
In at least one embodiment, the processing the three-dimensional point cloud data and the image in step 200 to obtain the spatial position and orientation information of the reference object further includes steps 210 to 240.
And step 210, carrying out joint calibration on the point cloud data acquisition equipment and the visual sensor to calculate coordinate conversion parameters.
Firstly, determining a calibration method of point cloud data acquisition equipment and a visual sensor, establishing a mathematical model based on a plane calibration object, calculating transformation matrix parameters between the point cloud data acquisition equipment and the visual sensor coordinates by using a least square method, performing error analysis after calculation, and adjusting after analysis. And finally determining coordinate conversion parameters of the point cloud data acquisition equipment and the visual sensor.
And step 220, modeling is carried out after the acquired three-dimensional point cloud data are processed, and the three-dimensional point cloud data of the reference object to be detected are screened out.
The method comprises the steps of carrying out a series of processing such as filtering, registering and dividing on the acquired point cloud data to obtain regional three-dimensional point cloud data including a reference object, carrying out feature extraction according to the geometric shape of the reference object, screening out a suspected reference object point cloud region in the regional point cloud data including the reference object by utilizing the geometric shape feature, removing the residual point cloud data, and carrying out accuracy detection on a screening result by using a classifier trained by an SVM mathematical model according to the geometric point cloud feature of the reference object.
And 230, calculating an area corresponding to the three-dimensional point cloud data of the reference object to be detected in the image by using the coordinate conversion parameters.
According to the three-dimensional point cloud area of the reference object determined in step 220, the corresponding reference object area in the image can be obtained by direct projection by using the coordinate conversion parameter between the point cloud data acquisition device and the visual sensor obtained by joint calibration in step 210.
And 240, performing decision fusion on the obtained three-dimensional point cloud data and the image, and comprehensively judging the spatial position and posture information of the reference object to be detected.
In order to fully utilize the three-dimensional point cloud data and the image information, in the embodiment, different classifiers are obtained by respectively training the three-dimensional point cloud data and the image to be detected independently, then decision fusion is performed on the results of the independent detection, comparison is performed, and when the detection results are inconsistent, the final detection result is judged according to a specific judgment criterion.
The specific judgment criterion is based on Bayes theory, firstly, the probability matrix is used for outputting and combining the detection results, the image information is divided into a reference object area (A) and a non-reference object area (B), the three-dimensional point cloud data is divided into reference object data (C) and non-reference object data (D), and four situations can occur: a + C; ② A + D; ③ B + C; fourthly, B + D; secondly, calculating the probability of the possible occurrence of the four situations according to a joint density function; secondly, training a classifier by using the characteristics of the three-dimensional point cloud data and detecting the target of the image information to obtain a target confidence coefficient, normalizing the obtained confidence coefficient, determining the prior probability of the two sensors, and determining a loss value with an incorrect judgment result by using a defined loss function; and finally determining a final judgment result according to a Bayes decision theory.
In at least one embodiment, step 300 is followed by step 310 after the adjustment of the electric wheelchair body.
And step 310, adjusting the seat of the electric wheelchair.
After the position and/or orientation of the electric wheelchair is calculated, the height of the electric wheelchair seat to be adjusted is calculated according to the height of the reference object,
and if the user wants to transfer the electric wheelchair to the target carrier, calculating the height of the electric wheelchair seat to be adjusted by taking the height of the hip contact of the user in the target carrier as a reference.
If a user wants to transfer the target carrier to the electric wheelchair, the reference object is the user and the target carrier, firstly, the target carrier where the user is located is determined, and when the target carrier is a soft surface, the difference between the height of the lowest point of the depression of the reference object and the height of the seat of the electric wheelchair is the adjustment amount of the seat of the electric wheelchair; when the target carrier is a hard surface, the height of the position of the target carrier contacted by the hip of the user in the reference object is the adjustment amount of the electric wheelchair seat. The electric wheelchair seat can be inclined by a small degree to assist a user to realize autonomous transfer.
The following is an example of an autonomous operation method of an electric wheelchair when a user transfers the electric wheelchair from the electric wheelchair to a target carrier.
A user wants to transfer to a sofa by taking the electric wheelchair, firstly, an instruction for transferring to the sofa is sent to the electric wheelchair, the electric wheelchair plans an optimal route according to a pre-established three-dimensional indoor map after receiving the instruction, and moves according to the planned optimal route to autonomously avoid obstacles; the method comprises the steps of determining a sofa as a reference object after the sofa is moved to the position near the sofa, acquiring information such as the height and the position of the sofa through point cloud data acquisition equipment and a visual sensor, calculating the direction, the angle and the distance of the electric wheelchair body, which are required by a user to transfer from the electric wheelchair to the sofa, needing to be adjusted, of the user through the electric wheelchair, calculating the height of the seat needing to be adjusted, adjusting the electric wheelchair according to a calculation result after the calculation is finished, and enabling the user to realize the autonomous transfer from the electric wheelchair to the sofa. After the user accomplishes the transfer action from carriers such as electronic wheelchair to sofa, bed, electronic wheelchair can independently plan the route, returns the standby space to prevent placing at will and occupy the indoor space or shelter from the sight, and can independently switch on in the standby space when the electric quantity is not enough, can also realize the electric quantity and be full of the back and independently cut off the power supply.
The following is an example of an autonomous operation method of an electric wheelchair when a user transfers from a target carrier to the electric wheelchair.
The method comprises the following steps that a user wants to transfer the electric wheelchair from a sofa to the electric wheelchair, firstly, an instruction for transferring the electric wheelchair to the sofa is sent to the electric wheelchair, the electric wheelchair plans an optimal route according to a pre-established three-dimensional indoor map after receiving the instruction, and moves according to the planned optimal route to autonomously avoid obstacles; after the electric wheelchair moves to the position near the sofa, the sofa and the user are determined to be reference objects, the information such as the height, the position, the posture and the like of the sofa and the user is obtained through the point cloud data acquisition equipment and the visual sensor, the electric wheelchair calculates the direction, the angle and the distance which are required to be adjusted by the user to transfer the electric wheelchair body from the electric wheelchair to the sofa, then the height which is required to be adjusted of the seat is calculated, and the adjustment is carried out according to the calculation result after the calculation is finished, so that the user can independently transfer the electric wheelchair to the sofa.
Referring to fig. 2, a system for autonomous operation based on an electric wheelchair disclosed in the present application will be described in detail, and as shown in fig. 2, the system disclosed in the present embodiment includes:
the autonomous navigation module is used for realizing autonomous movement of the electric wheelchair to the vicinity of a target carrier and automatically determining a reference object;
the point cloud data acquisition equipment is used for acquiring three-dimensional point cloud data and images of a reference object;
the reference object detection module is used for processing the three-dimensional point cloud data and the image to obtain spatial position and attitude information of a reference object;
the calculation and adjustment module is used for calculating the position adjustment amount and/or the orientation adjustment amount of the electric wheelchair body according to the spatial position posture information of the reference object, and adjusting the position and/or the orientation of the electric wheelchair body according to the calculation result until the position and/or the orientation meet the requirement that a user transfers between the electric wheelchair and the target carrier.
In at least one embodiment, the autonomous navigation module comprises:
the path planning unit is used for planning a moving path according to a pre-established three-dimensional indoor map after receiving the moving instruction;
the control unit is used for sending corresponding driving signals to the driving unit according to the planned moving path;
the driving unit is used for controlling the motor to operate according to the driving signal output by the control unit and driving the electric wheelchair body to move to the position near the target carrier through the motor;
the electric wheelchair obtains an environment point cloud of the surrounding environment by using a point cloud data acquisition device in the moving process, detects an obstacle based on the environment point cloud, replans a moving path according to the position of the obstacle and moves according to the newly planned moving path.
In at least one embodiment, the system further comprises a map creation module comprising:
the data acquisition unit is used for acquiring indoor three-dimensional point cloud data through point cloud data acquisition equipment of the electric wheelchair;
the data processing unit is used for filtering, registering and dividing the three-dimensional point cloud data collected by the data collecting unit to obtain a three-dimensional point cloud map;
the map conversion unit is used for converting the obtained three-dimensional point cloud map into a grid type three-dimensional indoor map.
In at least one embodiment, the system further comprises a map creation module comprising:
the joint calibration unit is used for carrying out joint calibration on the point cloud data acquisition equipment and the visual sensor to calculate coordinate conversion parameters;
the data screening unit is used for processing the acquired three-dimensional point cloud data and then modeling to screen out the three-dimensional point cloud data of the reference object to be detected;
the information conversion unit is used for calculating an area corresponding to the three-dimensional point cloud data of the reference object to be detected in the image by using the coordinate conversion parameters;
and the comprehensive judgment unit is used for performing decision fusion on the obtained three-dimensional point cloud data and the image and comprehensively judging the spatial position and posture information of the reference object to be detected.
In at least one embodiment, the target carrier and the user themselves are used as references when the user is transferred from the target carrier to the electric wheelchair;
when the user transfers the target carrier from the electric wheelchair, the target carrier is used as a reference object.
The division of the modules and units herein is only one division of logical functions, and other divisions may be possible in actual implementation, for example, a plurality of modules and/or units may be combined or integrated in another system. The modules and units described as separate parts may be physically separated or not. The components displayed as cells may or may not be physical cells, and may be located in a specific place or distributed in grid cells. Therefore, some or all of the units can be selected according to actual needs to implement the scheme of the embodiment.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present application should be covered within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.