CN210542119U - Intelligent wheelchair with obstacle avoidance and face recognition functions - Google Patents
Intelligent wheelchair with obstacle avoidance and face recognition functions Download PDFInfo
- Publication number
- CN210542119U CN210542119U CN201920641423.XU CN201920641423U CN210542119U CN 210542119 U CN210542119 U CN 210542119U CN 201920641423 U CN201920641423 U CN 201920641423U CN 210542119 U CN210542119 U CN 210542119U
- Authority
- CN
- China
- Prior art keywords
- wheelchair
- sensor
- computer
- rgb
- face recognition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Landscapes
- Traffic Control Systems (AREA)
Abstract
The utility model discloses an intelligent wheelchair with keep away barrier and face identification, including electronic wheelchair, install the on-vehicle computer on electronic wheelchair, install RGB-D sensor and the image acquisition module at wheelchair handrail front end, install the human-computer interaction module on the wheelchair handrail, install the GPS orientation module in wheelchair handrail below. The utility model scans the three-dimensional data around the wheelchair and in the moving direction through the RGB-D sensor on the electric wheelchair, synthesizes the three-dimensional image with the depth data, reminds the user to avoid the obstacle through the man-machine interaction module, and controls the brake of the wheelchair to avoid the obstacle urgently if necessary; the vehicle-mounted computer on the wheelchair can automatically recognize the human faces in the images and locally record and store the human faces, and in addition, a cloud server which can automatically record the people of the images and the position information of the shot images can be arranged for the user to manage and search according to time, people, places and the like, so that the user can search and review the images.
Description
Technical Field
The utility model relates to an intelligence wheelchair especially relates to an intelligence wheelchair with keep away barrier and face identification.
Background
With the continuous development of society, the economic strength is continuously improved, and the consumption level of the whole people is higher and higher. But the aging is more and more serious, and the demand of the intelligent wheelchair is more and more increased later. At present, the domestic research on the intelligent wheelchair is basically in the experimental research and design stage and does not have large-scale mass production, and the intelligent wheelchair still has a plurality of places needing improvement. Most intelligent wheelchairs only use a monocular sensor or an ultrasonic radar mode to identify obstacles. The monocular sensor can not identify or lose track when the target moves too fast, and the requirement on the computing power of a computer is high. Due to the principle limitation of the ultrasonic radar, the detection angle is small, large-area obstacles around the wheelchair cannot be identified, and the obstacles cannot be classified and visualized, so that the system cannot accurately avoid the obstacles and identify the human face.
By looking up the literature on the chinese Hopkinson Web, two general categories can be distinguished.
The first intelligent wheelchair utilizes a monocular camera to acquire image data, and detects obstacles through the size change of an object between two continuous frames. Or the obstacle is positioned by using a deformation grid method, the obstacle can be gradually deformed along with the change of the object in the scene, and the method has higher efficiency than the former method, but cannot carry out specific classification and identification on the obstacle.
The second intelligent wheelchair mainly adopts a transit time detection method of an ultrasonic ranging method, namely under the condition that the sound velocity is known, the distance is obtained by measuring the time of ultrasonic echo. The information of the obstacle is detected by using 3 or more ultrasonic sensors, and is generally installed right in front of the wheelchair, left in front of the wheelchair, and right in front of the wheelchair. When the ultrasonic sensor works, the single chip microcomputer is used for gating the multi-path analog switches through the two signal lines to enable the multi-path analog switches to work sequentially, and then the multi-path analog switches are used for controlling the on-off of each path of the ultrasonic sensor. The next transducer starts to emit ultrasonic waves after the previous transducer receives the reflected wave. The use of ultrasound for feedback of the position of obstacles has some drawbacks. The direct reflection type ultrasonic sensor cannot reliably detect an obstacle positioned at the front end of the ultrasonic transducer, the ultrasonic transducer and the area of the starting point of the detection range have certain blind areas, the problem of multiple reflection occurs when the wall corner is measured, sound waves are received by the sensor after multiple reflection, the actual measured value is not a real distance value, and noise with similar frequency generated by the surrounding environment interferes the measured result.
Disclosure of Invention
An object of the utility model is to overcome above-mentioned prior art not enough, provide one kind can the barrier during use, take face identification's intelligent wheelchair.
In order to realize the purpose, the utility model discloses a technical scheme is: the utility model provides an intelligence wheelchair with keep away barrier and face identification, includes electronic wheelchair, its characterized in that: the electric wheelchair comprises an electric wheelchair body, and is characterized by further comprising an on-board computer, an RGB-D sensor, a GPS positioning module, an image acquisition module and a human-computer interaction module, wherein the on-board computer is installed below a seat plate of the electric wheelchair and connected with a controller of the electric wheelchair, the RGB-D sensor and the image acquisition module are installed at the front end of a handrail of the electric wheelchair respectively, the GPS positioning module is installed on the bottom surface of the handrail of the electric wheelchair, and the RGB-D sensor, the GPS positioning module and the image acquisition module are connected with the on-board computer respectively.
Specifically, the vehicle-mounted computer is mounted below a seat plate or on a backrest plate of the electric wheelchair.
Further, the RGB-D sensor is a Kinect sensor, the image acquisition module is a high-definition wide-angle camera, and the Kinect sensor and the high-definition wide-angle camera are respectively connected with NVIDIA TK1 through USB connecting wires. The human-computer interaction module is a tablet computer and is connected with NVIDIA TK1 through a network.
Further, the utility model discloses still include a cloud ware, cloud ware with the on-vehicle computer passes through the network and links to each other.
The utility model has the advantages that: scanning three-dimensional data around the wheelchair and in the moving direction by an RGB-D sensor arranged on the electric wheelchair, synthesizing a three-dimensional image with depth data, reminding a user to avoid obstacles by a human interaction module (in an image or voice mode), and automatically controlling the brake of the wheelchair by a system when necessary to avoid obstacles urgently; the image acquisition module on the electric wheelchair can shoot the surrounding environment of the wheelchair, and the vehicle-mounted computer on the wheelchair automatically analyzes the acquired images, so that the images can be automatically identified and recorded if people meet the images. The result of identifying people can prompt the user through a human-computer interaction module on the wheelchair, for example, a display or voice mode, and the identified image can be automatically filed and uploaded to a cloud server for the user to manage and search according to time, people, places and the like, so that the user can search and review the image.
Drawings
The present invention will be described in further detail with reference to the accompanying drawings.
Fig. 1 is a schematic structural diagram of the present invention.
The system comprises a wheelchair 1, a vehicle-mounted computer 2, an RGB-D sensor 3, a GPS positioning module 4, an image acquisition module 5, a human-computer interaction module 6 and a human-computer interaction module.
Detailed Description
As shown in figure 1, the utility model relates to an intelligent wheelchair with keep away barrier and face identification, including electronic wheelchair 1 to and on-vehicle computer 2, RGB-D sensor 3, GPS orientation module 4, image acquisition module 5 and human-computer interaction module 6. The RGB-D sensor 3 and the image acquisition module 5 are respectively installed at the front end of the armrest of the electric wheelchair, and the GPS positioning module 4 is installed on the bottom surface of the armrest of the electric wheelchair. In the embodiment, the vehicle-mounted computer 2 is arranged on a backboard of the electric wheelchair and is connected with the controller of the electric wheelchair 1; of course, the cycle computer 2 can also be mounted under the seat plate of the electric wheelchair.
The vehicle mount computer 2 developed a component for NVIDIA TK 1. The RGB-D sensor 3 is a Kinect sensor, the image acquisition module 5 is a high-definition wide-angle camera, and the Kinect sensor and the high-definition wide-angle camera are respectively connected with NVIDIA TK1 through USB connecting wires. The human-computer interaction module 6 is a tablet computer and is connected with NVIDIA TK1 through a network.
The NVIDIA TK1 development component is a full-function NVIDIA CUDA platform, and can rapidly develop and deploy a calculation-intensive system used in the fields of computer vision, robotics, medicine and the like. NVIDIA provides BSPs and software packages including CUDA, OpenGL 4.4, and NVIDIA VisionWorks. The user can rely on complete development and dynamic analysis tool and develop camera and other peripheral equipments, NVIDIA TK1 development subassembly has included System On Chip (SOC), NVIDIA Kepler GPU of 192 CUDA cores, NVIDIA (4-Plus-1), four-core ARM Cortex-A15 CPU, 2 GB operation memory, 16 GB eMMC storage space, gigabit Ethernet, USB 3.0 SD/MMC, miniPCIE HDMI 1.4 SATA line output/microphone input, RS232 serial port, expansion port is used for extra display, general IO and high bandwidth camera interface etc., Jetson TK1 supports Linux, be fit for application and system development in each industry. The requirements of the intelligent wheelchair on image processing and wheelchair control can be met. An external interface arranged on the TK1 can be conveniently connected with other equipment on the intelligent wheelchair including a Kinect sensor.
The RGB-D sensor on the wheelchair armrest adopts a Microsoft Kinect sensor, the Kinect has three lenses, and the middle lens is an RGB color camera and is used for collecting color images. The left and right lenses are 3D structured light depth sensors composed of an infrared emitter and an infrared CMOS camera, and are used for collecting depth data, namely the distance from an object in a scene to the camera. Kinect uses the Time of Flight (TOF) technique to obtain three-dimensional depth data of 640 x 480 resolution. Therefore, three-dimensional space data in front of the intelligent wheelchair can be reconstructed, and whether obstacles exist on the running route of the vehicle or not can be judged by analyzing the data.
The Kinect transmits the image depth data to the vehicle-mounted computer NVIDIA TK1 through a USB connecting line, NVIDIATK1 calculates the position and distance of an obstacle through the depth data returned by the Kinect, the depth data is displayed on a tablet computer through a color three-dimensional graph with depth by using a point cloud technology (PCL), and the obstacle which is possibly harmful to driving safety is shown on a display through warning color (red) after calculation, and is prompted through voice on the tablet computer.
The image acquisition module 5 adopts a high-definition wide-angle camera, can shoot 1920 × 1080 resolution, has 30 frames of sampling frequency, and is connected with the NVIDIA TK1 through a USB connecting line.
NVIDIA TK1 is connected with a Baidu cloud server in real time through a 4G network. The NVIDIA TK1 analyzes the color image data acquired by the image acquisition module in real time, a face image existing in the image is extracted, once the face existing in the image is found, the face image is sent to the Baidu cloud server through the 4G network for recognition processing, the current face data and the existing face data set are compared, the comparison and recognition results are fed back to the vehicle-mounted computer NVIDIA TK1, and the user is prompted through a display screen or a voice mode.
The GPS module 4 is connected with NVIDIA TK1 through a USB connecting line, and records the current GPS position in real time.
Further, the utility model discloses still include a cloud ware, cloud ware with the on-vehicle computer passes through the network and links to each other.
The utility model discloses intelligence wheelchair with keep away barrier and face identification is around the scanning wheelchair and the ascending three-dimensional data of moving direction through RGB-D sensor at the walking in-process, reminds the user through human-computer interaction module (image or speech mode) to keep away the barrier, and the brake of system automatic control wheelchair promptly keeps away the barrier when necessary. Meanwhile, the image acquisition module on the electric wheelchair can shoot the surrounding environment of the wheelchair in real time, and the image can be identified and recorded if people meet the image. The recognition result can prompt the user through the display of a human-computer interaction module on the wheelchair or a voice mode. And the identified image and the GPS position information are recorded and filed through the cloud server, so that the user can conveniently inquire at any time.
The above description is only for the purpose of illustrating the technical solutions of the present invention, and the simple modification or equivalent replacement of the technical solutions of the present invention by those of ordinary skill in the art does not depart from the spirit and scope of the technical solutions of the present invention.
Claims (7)
1. The utility model provides an intelligence wheelchair with keep away barrier and face identification, includes electronic wheelchair, its characterized in that: the electric wheelchair comprises an electric wheelchair body, and is characterized by further comprising an on-board computer, an RGB-D sensor, a GPS positioning module, an image acquisition module and a human-computer interaction module, wherein the on-board computer is installed on the electric wheelchair body and connected with a controller of the electric wheelchair body, the RGB-D sensor and the image acquisition module are installed at the front end of a handrail of the electric wheelchair body respectively, the GPS positioning module is installed on the bottom surface of the handrail of the electric wheelchair body, and the RGB-D sensor, the GPS positioning module and the image acquisition module are connected with the on-board computer respectively.
2. The intelligent wheelchair with obstacle avoidance and face recognition as claimed in claim 1, wherein: the on-board computer develops a component for NVIDIA TK 1.
3. The intelligent wheelchair with obstacle avoidance and face recognition as claimed in claim 1, wherein: the RGB-D sensor is a Kinect sensor and is connected with NVIDIA TK1 through a USB connecting line.
4. The intelligent wheelchair with obstacle avoidance and face recognition as claimed in claim 1, wherein: the image acquisition module is a high-definition wide-angle camera and is connected with NVIDIA TK1 through a USB connecting wire.
5. The intelligent wheelchair with obstacle avoidance and face recognition as claimed in claim 1, wherein: the human-computer interaction module is a tablet computer and is connected with NVIDIA TK1 through a network.
6. The intelligent wheelchair with obstacle avoidance and face recognition as claimed in any one of claims 1 to 5, wherein: the vehicle-mounted computer system further comprises a cloud server, and the cloud server is connected with the vehicle-mounted computer through a network.
7. The intelligent wheelchair with obstacle avoidance and face recognition as claimed in claim 6, wherein: the vehicle-mounted computer is arranged below a seat plate or on a backrest plate of the electric wheelchair.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201920641423.XU CN210542119U (en) | 2019-05-07 | 2019-05-07 | Intelligent wheelchair with obstacle avoidance and face recognition functions |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201920641423.XU CN210542119U (en) | 2019-05-07 | 2019-05-07 | Intelligent wheelchair with obstacle avoidance and face recognition functions |
Publications (1)
Publication Number | Publication Date |
---|---|
CN210542119U true CN210542119U (en) | 2020-05-19 |
Family
ID=70664193
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201920641423.XU Expired - Fee Related CN210542119U (en) | 2019-05-07 | 2019-05-07 | Intelligent wheelchair with obstacle avoidance and face recognition functions |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN210542119U (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112089559A (en) * | 2020-08-18 | 2020-12-18 | 西安交通大学 | Auxiliary standing device and method based on indoor positioning and artificial intelligence |
CN114429497A (en) * | 2020-10-14 | 2022-05-03 | 西北农林科技大学 | Living body Qinchuan cattle body ruler measuring method based on 3D camera |
CN115154080A (en) * | 2022-07-07 | 2022-10-11 | 广东职业技术学院 | Anti-collision system and method for electric wheelchair |
-
2019
- 2019-05-07 CN CN201920641423.XU patent/CN210542119U/en not_active Expired - Fee Related
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112089559A (en) * | 2020-08-18 | 2020-12-18 | 西安交通大学 | Auxiliary standing device and method based on indoor positioning and artificial intelligence |
CN112089559B (en) * | 2020-08-18 | 2021-06-01 | 西安交通大学 | Auxiliary standing device and method based on indoor positioning and artificial intelligence |
CN114429497A (en) * | 2020-10-14 | 2022-05-03 | 西北农林科技大学 | Living body Qinchuan cattle body ruler measuring method based on 3D camera |
CN115154080A (en) * | 2022-07-07 | 2022-10-11 | 广东职业技术学院 | Anti-collision system and method for electric wheelchair |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN210542119U (en) | Intelligent wheelchair with obstacle avoidance and face recognition functions | |
KR102614323B1 (en) | Create a 3D map of a scene using passive and active measurements | |
KR102109941B1 (en) | Method and Apparatus for Vehicle Detection Using Lidar Sensor and Camera | |
WO2019179417A1 (en) | Data fusion method and related device | |
CN110582778B (en) | Embedded motor vehicle awareness with machine learning classification of sensor data | |
JP5748920B2 (en) | How to display around the vehicle | |
KR102670610B1 (en) | Robot for airport and method thereof | |
EP3872688A1 (en) | Obstacle identification method and device, storage medium, and electronic device | |
EP3709134A1 (en) | Tool and method for annotating a human pose in 3d point cloud data | |
CN103455144A (en) | Vehicle-mounted man-machine interaction system and method | |
CN111753609A (en) | Target identification method and device and camera | |
CN203365452U (en) | Vehicle speed detection system based on binocular stereoscopic vision | |
CN114442101B (en) | Vehicle navigation method, device, equipment and medium based on imaging millimeter wave radar | |
CN112330746A (en) | Mobile chassis obstacle detection method based on TX2 | |
CN101234601A (en) | Automobile cruise control method based on monocular vision and implement system thereof | |
CN108608440A (en) | Service robot and service robot system | |
US10114468B2 (en) | Method for evaluating gestures | |
Gazis et al. | Examining the sensors that enable self-driving vehicles | |
CN211527320U (en) | Automobile-used laser figure surveys barrier device | |
US11346944B2 (en) | Classification of static and dynamic objects | |
CN114359865A (en) | Obstacle detection method and related device | |
CA3074628A1 (en) | Method and device for optical distance measurement | |
CN117197779A (en) | Track traffic foreign matter detection method, device and system based on binocular vision | |
JP6972798B2 (en) | Information processing device, image pickup device, device control system, mobile body, information processing method, and program | |
JPS62231190A (en) | Collision warning device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200519 Termination date: 20210507 |