CN211273455U

CN211273455U - Intelligent ball picking machine vehicle based on automatic image processing

Info

Publication number: CN211273455U
Application number: CN201922263061.8U
Authority: CN
Inventors: 张婵; 罗佳
Original assignee: Guangdong Industry Technical College
Current assignee: Guangdong Industry Technical College
Priority date: 2019-12-16
Filing date: 2019-12-16
Publication date: 2020-08-18
Anticipated expiration: 2029-12-16

Abstract

The patent discloses an intelligent ball-picking machine vehicle based on automatic image processing, which comprises a vehicle body, ball-picking mechanical arms and wheels at the front part of the vehicle body, and also comprises a driving expansion board, an ultrasonic device, an infrared sensor, a camera and a battery which are respectively and electrically connected with a raspberry group main board, the driving expansion board is connected with a motor driving device and respectively controls and drives the ball picking mechanical arm and the wheel, the processor of the raspberry pi main board carries out intelligent calculation and identification on the images acquired by the camera by calling the OpenCV image library, controls the movement of the vehicle body and the movement of the ball-picking mechanical arm to pick the balls, the distance between the vehicle body and a target object is measured through the ultrasonic device, the vehicle body is controlled to move, the position of a small ball and an obstacle are automatically identified, the table tennis ball or the tennis ball is flexibly and movably grabbed, manual ball picking work is replaced, the structure is simple, the cost is low, and the device is suitable for large-scale sports places.

Description

Intelligent ball picking machine vehicle based on automatic image processing

Technical Field

The patent relates to a robot that ball was picked up to intelligence.

Background

In recent years, with the rapid development of mobile robots, in various ball facilities nowadays, a pitching machine is scaled and ranked, however, intelligent ball picking machines are few, such as table tennis balls and tennis balls, and although ball picking machines exist in the market, the intelligent ball picking machines generally have some obvious disadvantages. Firstly, manual control is needed; secondly, the ball picker in the current market is large in size, is difficult to work indoors after weighing more than 50 jin, greatly reduces the flexibility and is heavy; moreover, the current ball picking machine is generally expensive in manufacturing cost and high in maintenance cost, is only suitable for large-scale sports places, is small in target group and is not suitable for mass consumption.

Compared with the traditional computer, the raspberry pi mainboard serving as a microcomputer is only short of a display screen and a keyboard, and is widely applied to image and word processing. Raspberry group 3B + mainboard possess powerful treater and can carry out real-time processing to the image that the camera was gathered, accomplish and gather and discernment, and the USB port can conveniently be connected LINUX and exempt from to drive the camera, has independent 40 GPIO pins, can be when accomplishing image processing output motor, the control signal of arm, the HDMI interface can insert the convenient image display of display, has 4 USB interfaces and ethernet wireless network card. The raspberry group can run Python language for compiling, a Python-GPIO function library is built in the raspberry group, a bottom-layer IO control driver does not need to be independently compiled when a GPIO pin is used for outputting a control signal, and the Python-GPIO function library can be directly used for driving GPIO. The raspberry group has high operation speed and low price, and is a good control platform.

Disclosure of Invention

The utility model provides a ball machine car is picked up to intelligence based on automatic image processing, but automatic identification bobble position and barrier move in a flexible way and snatch the bobble, replace artifical completion to pick up ball work, simple structure is with low costs, is fit for large-scale motion place and uses.

Ball machine tool car is picked up to intelligence based on automatic image processing, including automobile body, the anterior ball arm of picking up of automobile body, wheel, still include respectively with raspberry group mainboard electric connection's drive expansion board, ultrasonic device, infrared sensor, camera, battery, the drive expansion board is connected motor drive, the drive of controlling respectively ball arm and wheel are picked up to the drive, be provided with the image storehouse in the memory of raspberry group mainboard, through intelligent recognition the image that the camera was gathered to measure the distance between automobile body and the target object through ultrasonic device, control vehicle removes and picks up the ball arm motion and snatchs the bobble.

Further, raspberry group mainboard is raspberry group 3B + mainboard, sets up in the automobile body, through first GPIO interface connection drive expansion board, second GPIO interface connection ultrasonic device, third GPIO interface connection infrared sensor, USB interface connection camera.

Further, an OpenCV image library is arranged in a memory of the raspberry group main board, and a processor of the raspberry group main board carries out intelligent calculation and recognition by calling the OpenCV image library, and controls the car body to move and the ball picking mechanical arm to move to pick the balls. The raspberry pi main board is provided with an Ethernet card and is connected with a mobile phone through a WIFI interface to transmit signals.

Further, the motor driving device comprises a transmission motor, a steering motor and a ball picking mechanical arm motion motor, wherein the transmission motor and the steering motor respectively control the trolley to move and steer; the ball picking mechanical arm moving motor is connected with the ball picking mechanical arm through a driver and used for controlling the clamping of the ball body.

This patent is based on automatic image processing's intelligence and is picked up ball machine car and bring following advantage:

1. the intelligent ball picking robot vehicle integrates an automation technology and an image recognition technology, and the raspberry group main board is used as a core platform by applying the image processing technology, so that the action of the robot can be controlled while the image processing is finished, and the system is simple in structure. The raspberry pi is used as a platform, the OpenCV is used as an image processing tool, the image processing process is simplified, and the recognition rate is improved. The robot has the advantages of simple mechanical structure, high flexibility, strong maneuverability, portability, easy maintenance and low cost, and can autonomously identify and grab table tennis balls and tennis balls. The mobile phone can also be used for controlling the movement of the robot, so that the robot adapts to the real environment to the maximum extent.

2. The cooperativity and the operability are good. The raspberry pi main board is used as a core platform for image processing, an OpenCV image library is called by a Python interface to complete image recognition, the image recognition can be directly combined with an execution mechanism, table tennis balls can be automatically found, the shortest path can be determined, the functions of moving obstacle avoidance and cooperatively picking up the table tennis balls are achieved, a series of actions of recognizing the table tennis balls, avoiding obstacles and picking up the table tennis balls are completed, and the intelligent equipment is provided. The raspberry pi main board is provided with an Ethernet card, and can also wirelessly transmit data in real time through a mobile phone, so that automatic operation and remote control command operation are integrated.

3. The intelligent ball picking machine car is simple in structure, low in cost and high in practicability, replaces manual work to finish ball picking work, saves physical force wasted by athletes in training or competition due to ball picking, and improves sports experience of the athletes.

Drawings

FIG. 1 is a schematic structural diagram of an intelligent ball-picking robot vehicle based on automatic image processing provided by an embodiment;

fig. 2 is a circuit connection schematic diagram of a raspberry pi main board of the intelligent ball picking machine vehicle.

Detailed Description

To make the objects, technical solutions and advantages of the present patent more clear, the present patent will be further described in detail with reference to the accompanying drawings.

As shown in fig. 1, 2, the ball machine car is picked up to intelligence based on automatic image processing that this patent provided, including automobile body 1, the anterior ball arm 2, the wheel 3 of picking up of automobile body, still including sending 4 electric connection's drive expansion board 5, ultrasonic device 6, infrared sensor 7, camera 8, battery 9 with the raspberry respectively, motor drive 55 is connected to the drive expansion board, controls the drive respectively ball arm and wheel pick up, the raspberry is sent the mainboard and is provided with the image storehouse, through intelligent recognition the image that the camera was gathered to measure the distance between automobile body and the target object through ultrasonic device, control automobile body removes and picks up ball arm motion and snatchs the bobble. The raspberry group mainboard is a raspberry group 3B + mainboard, and is connected with the driving expansion board through a first GPIO interface, a second GPIO interface is connected with the ultrasonic device, a third GPIO interface is connected with the infrared sensor, and a USB interface is connected with the camera. The camera is installed and is scanned the place in the place ahead in automobile body top, discernment barrier and spheroid, and transmission signal sends the mainboard for the raspberry, judges whether snatch the spheroid. The ultrasonic device is arranged in the middle of the front part of the vehicle body, and sends out ultrasonic detection front obstacles transmission signals to the raspberry group mainboard, so that the trolley is automatically controlled to avoid. The two infrared sensors are arranged on two sides below the ultrasonic device and assist in detecting obstacles in front and transmitting signals to the raspberry group main board, if the infrared detector on the left detects the obstacles, the trolley is automatically controlled to turn to the right, and if the infrared detector on the right detects the obstacles, the trolley is automatically controlled to turn to the left.

The system develops languages: python, Java, development system based: the system comprises a Linux drive-free USB camera, wherein an OpenCV image library is arranged in a memory of a raspberry serving main board, a camera is started to collect a frame of picture of a target object, a processor of the raspberry serving main board calls the OpenCV image library to perform intelligent calculation and identification, an OpenCV default read image is an RGB channel and is converted into a gray picture, Hough edge detection is performed to return the edge number of the target object so as to complete identification of the target object, the distance from a recognized sphere to a trolley is determined according to the maximum radius and the minimum radius of identification parameters, and a vehicle body is controlled to move and a sphere picking mechanical arm is controlled to move to pick a sphere.

The raspberry pi main board is provided with an Ethernet card, is connected with a mobile phone through a WIFI interface to transmit signals, and can remotely control the trolley to stop working or start working. The motor driving device comprises a transmission motor, a steering motor and a ball picking mechanical arm motion motor, wherein the transmission motor and the steering motor respectively control the trolley to move and steer; the ball picking mechanical arm moving motor is connected with the ball picking mechanical arm through a driver and used for controlling the clamping of the ball body.

The intelligent ball picking machine vehicle identifies balls based on an OpenCV computer vision library, a raspberry group 3B + main board is used as a core platform for image processing, an ultrasonic sensor is used for obstacle avoidance, the ultrasonic obstacle avoidance refers to a method for obtaining the distance between an obstacle and the own party through ultrasonic ranging, and the approximate direction and distance of a target are determined. The ultrasonic generator T sends out an ultrasonic signal at a certain moment, and the ultrasonic signal is reflected back to be received by the ultrasonic receiver R after meeting a measured object. Assuming that the received distance data is a, if a < > is 20cm, judging that an obstacle exists in front, and backing the trolley to perform left-turn or right-turn operation; if a is larger than 20cm, it is judged that there is no obstacle in front and the vehicle moves forward. When the table tennis ball detection device is operated, a camera and an infrared sensor of the trolley are opened in real time to scan a field in front, the table tennis ball is identified, when the table tennis ball is identified, the distance between the trolley and the table tennis ball is adjusted according to the set radius range, for example, the distance is 5 cm, the table tennis ball is picked up and sent to the basket by the mechanical arm, and then the table tennis ball is continuously walked to detect the table tennis ball.

The working principle of the intelligent ball picking machine vehicle based on automatic image processing is described as follows: firstly, after the raspberry pi 3B + main board starts the ultrasonic ranging device to measure the distance between the intelligent ball picking vehicle and a target object, the driving motor rotates forwards or leftwards and rightwards, and the infrared sensor assists in detecting the obstacle. And when the mobile terminal runs for a certain distance, the mobile terminal stops moving, starts a camera module to collect a frame of image, and returns the result to the controller of the raspberry pi 3B + mainboard after image processing. And when the correct recognition result of the small balls is obtained, the mechanical arm is driven to complete the grabbing of the target ball. When the target object cannot be identified correctly, after the positions of the camera and the target object are automatically calculated, the intelligent ball picking vehicle is identified again and operated until a correct result is generated to grab the target ball.

The intelligent ball picking machine vehicle integrates an automation technology and an image recognition technology, can automatically recognize the positions of the balls and the obstacles, flexibly moves and picks the balls, replaces manual work to finish ball picking work, has a simple structure and low cost, can conveniently upgrade a control program and improve functions, and is suitable for large-scale sports places.

Claims

1. The intelligent ball picking machine vehicle based on automatic image processing comprises a vehicle body, ball picking mechanical arms and wheels, and is characterized by further comprising a driving expansion board, an ultrasonic device, an infrared sensor, a camera and a battery, wherein the driving expansion board is electrically connected with a raspberry sending main board, the driving expansion board is connected with a motor driving device and controls and drives the ball picking mechanical arms and the wheels respectively, an image library is arranged in a memory of the raspberry sending main board, images collected by the camera are intelligently identified, the distance between the vehicle body and a target object is measured through the ultrasonic device, and the vehicle is controlled to move and the ball picking mechanical arms are controlled to pick balls.

2. The intelligent robot car of picking up ball based on automatic image processing of claim 1, characterized in that, the raspberry pi mainboard is raspberry pi 3B + mainboard, through first GPIO interface connection drive expansion board, second GPIO interface connection ultrasonic device, third GPIO interface connection infrared sensor, USB interface connection camera.

3. The intelligent robot picking vehicle based on automatic image processing according to claim 1, wherein an OpenCV image library is arranged in a memory of the raspberry pi, and a processor of the raspberry pi performs intelligent calculation and recognition by calling the OpenCV image library, and controls a vehicle body to move and a ball picking mechanical arm to move to pick a ball.

4. The intelligent ball picking machine vehicle based on automatic image processing as claimed in claim 1, wherein the motor driving device comprises a transmission motor, a steering motor and a ball picking mechanical arm motion motor, and the transmission motor and the steering motor respectively control the movement and the steering of the trolley; the ball picking mechanical arm moving motor is connected with the ball picking mechanical arm through a driver and used for controlling the clamping of the ball body.

5. The intelligent robot picking vehicle based on automatic image processing is characterized in that the raspberry pi main board is provided with an Ethernet card, and is connected with a mobile phone through a WIFI interface to transmit signals.