OpenCV物体跟踪树莓派视觉小车实现过程学习

物体跟踪效果展示

 

过程：

一、初始化

def motor_init():
    global l_motor, r_motor
    l_motor= gpio.pwm(l_motor,100)
    r_motor = gpio.pwm(r_motor,100)
    l_motor.start(0)
    r_motor.start(0) 
def direction_init():
    gpio.setup(left_back,gpio.out)
    gpio.setup(left_front,gpio.out)
    gpio.setup(l_motor,gpio.out)
    
    gpio.setup(right_front,gpio.out)
    gpio.setup(right_back,gpio.out)
    gpio.setup(r_motor,gpio.out)  
def servo_init():
    global pwm_servo
    pwm_servo=adafruit_pca9685.pca9685()
def init():
    gpio.setwarnings(false) 
    gpio.setmode(gpio.bcm)
    direction_init()
    servo_init()
    motor_init()

二、运动控制函数

def front(speed):
    l_motor.changedutycycle(speed)
    gpio.output(left_front,1)   #left_front
    gpio.output(left_back,0)    #left_back
    r_motor.changedutycycle(speed)
    gpio.output(right_front,1)  #right_front
    gpio.output(right_back,0)   #right_back      
def back(speed):
    l_motor.changedutycycle(speed)
    gpio.output(left_front,0)   #left_front
    gpio.output(left_back,1)    #left_back 
    r_motor.changedutycycle(speed)
    gpio.output(right_front,0)  #right_front
    gpio.output(right_back,1)   #right_back 
def left(speed):
    l_motor.changedutycycle(speed)
    gpio.output(left_front,0)   #left_front
    gpio.output(left_back,1)    #left_back
    r_motor.changedutycycle(speed)
    gpio.output(right_front,1)  #right_front
    gpio.output(right_back,0)   #right_back
def right(speed):
    l_motor.changedutycycle(speed)
    gpio.output(left_front,1)   #left_front
    gpio.output(left_back,0)    #left_back 
    r_motor.changedutycycle(speed)
    gpio.output(right_front,0)  #right_front
    gpio.output(right_back,1)   #right_back 
def stop():
    l_motor.changedutycycle(0)
    gpio.output(left_front,0)   #left_front
    gpio.output(left_back,0)    #left_back
    r_motor.changedutycycle(0)
    gpio.output(right_front,0)  #right_front
    gpio.output(right_back,0)   #right_back

三、舵机角度控制

def set_servo_angle(channel,angle):
    angle=4096*((angle*11)+500)/20000
    pwm_servo.set_pwm_freq(50)                #frequency==50hz (servo)
    pwm_servo.set_pwm(channel,0,int(angle))

set_servo_angle(4, 110)     #top servo     lengthwise
    #0:back    180:front    
    set_servo_angle(5, 90)     #bottom servo  crosswise
    #0:left    180:right  

上面的（4）：是顶部的舵机（摄像头上下摆动的那个舵机）

下面的（5）：是底部的舵机（摄像头左右摆动的那个舵机）

四、摄像头&&图像处理

# 1 image process
        img, contours = image_processing()

width, height = 160, 120
    camera = cv2.videocapture(0)
    camera.set(3,width) 
    camera.set(4,height) 

1、打开摄像头

打开摄像头，并设置窗口大小。

设置小窗口的原因： 小窗口实时性比较好。

# capture the frames
    ret, frame = camera.read()

2、把图像转换为灰度图

# to gray
gray = cv2.cvtcolor(frame, cv2.color_bgr2gray)
cv2.imshow('gray',gray)

3、 高斯滤波（去噪）

# gausi blur
    blur = cv2.gaussianblur(gray,(5,5),0)

4、亮度增强

#brighten
    blur = cv2.convertscaleabs(blur, none, 1.5, 30)

5、转换为二进制

#to binary
    ret,binary = cv2.threshold(blur,150,255,cv2.thresh_binary_inv)
    cv2.imshow('binary',binary)

6、闭运算处理

#close
    kernel = cv2.getstructuringelement(cv2.morph_rect, (17,17))
    close = cv2.morphologyex(binary, cv2.morph_close, kernel)
    cv2.imshow('close',close)

7、获取轮廓

#get contours
    binary_c,contours,hierarchy = cv2.findcontours(close, 1, cv2.chain_approx_none)
    cv2.drawcontours(image, contours, -1, (255,0,255), 2)
    cv2.imshow('image', image)

代码

def image_processing():
    # capture the frames
    ret, frame = camera.read()
    # crop the image
    image = frame
    cv2.imshow('frame',frame)
    # to gray
    gray = cv2.cvtcolor(frame, cv2.color_bgr2gray)
    cv2.imshow('gray',gray)
    # gausi blur
    blur = cv2.gaussianblur(gray,(5,5),0)
    #brighten
    blur = cv2.convertscaleabs(blur, none, 1.5, 30)
    #to binary
    ret,binary = cv2.threshold(blur,150,255,cv2.thresh_binary_inv)
    cv2.imshow('binary',binary)
    #close
    kernel = cv2.getstructuringelement(cv2.morph_rect, (17,17))
    close = cv2.morphologyex(binary, cv2.morph_close, kernel)
    cv2.imshow('close',close)
    #get contours
    binary_c,contours,hierarchy = cv2.findcontours(close, 1, cv2.chain_approx_none)
    cv2.drawcontours(image, contours, -1, (255,0,255), 2)
    cv2.imshow('image', image)
    return frame, contours

五、获取最大轮廓坐标

由于有可能出现多个物体，我们这里只识别最大的物体（深度学习可以搞分类，还没学到这，学到了再做），得到它的坐标。

# 2 get coordinates
        x, y = get_coord(img, contours)

def get_coord(img, contours):
    image = img.copy()
    try:
        contour = max(contours, key=cv2.contourarea)
        cv2.drawcontours(image, contour, -1, (255,0,255), 2)
        cv2.imshow('new_frame', image)
        # get coord
        m = cv2.moments(contour)
        x = int(m['m10']/m['m00'])
        y = int(m['m01']/m['m00'])
        print(x, y) 
        return x,y
        
    except:
        print 'no objects'
        return 0,0

返回最大轮廓的坐标：

六、运动

根据反馈回来的坐标，判断它的位置，进行运动。

# 3 move
        move(x,y)

1、没有识别到轮廓（静止）

    if x==0 and y==0:
        stop()

2、向前走

识别到物体，且在正*（中间1/2区域），让物体向前走。

#go ahead
    elif width/4 <x and x<(width-width/4):
        front(70)

3、向左转

物体在左边1/4区域。

#left
    elif x < width/4:
        left(50)

4、向右转

物体在右边1/4区域。

#right
    elif x > (width-width/4):
        right(50)

代码

def move(x,y):
    global second
    #stop
    if x==0 and y==0:
        stop()
    #go ahead
    elif width/4 <x and x<(width-width/4):
        front(70)
    #left
    elif x < width/4:
        left(50)
    #right
    elif x > (width-width/4):
        right(50)

总代码

#object tracking
import  rpi.gpio as gpio
import time
import adafruit_pca9685
import numpy as np
import cv2
second = 0 
width, height = 160, 120
camera = cv2.videocapture(0)
camera.set(3,width) 
camera.set(4,height) 
l_motor = 18
left_front   =  22
left_back   =  27
r_motor = 23
right_front   = 25
right_back  =  24 
def motor_init():
    global l_motor, r_motor
    l_motor= gpio.pwm(l_motor,100)
    r_motor = gpio.pwm(r_motor,100)
    l_motor.start(0)
    r_motor.start(0) 
 def direction_init():
    gpio.setup(left_back,gpio.out)
    gpio.setup(left_front,gpio.out)
    gpio.setup(l_motor,gpio.out)    
    gpio.setup(right_front,gpio.out)
    gpio.setup(right_back,gpio.out)
    gpio.setup(r_motor,gpio.out) 
def servo_init():
    global pwm_servo
    pwm_servo=adafruit_pca9685.pca9685()
def init():
    gpio.setwarnings(false) 
    gpio.setmode(gpio.bcm)
    direction_init()
    servo_init()
    motor_init()
def front(speed):
    l_motor.changedutycycle(speed)
    gpio.output(left_front,1)   #left_front
    gpio.output(left_back,0)    #left_back
    r_motor.changedutycycle(speed)
    gpio.output(right_front,1)  #right_front
    gpio.output(right_back,0)   #right_back   
def back(speed):
    l_motor.changedutycycle(speed)
    gpio.output(left_front,0)   #left_front
    gpio.output(left_back,1)    #left_back 
    r_motor.changedutycycle(speed)
    gpio.output(right_front,0)  #right_front
    gpio.output(right_back,1)   #right_back 
def left(speed):
    l_motor.changedutycycle(speed)
    gpio.output(left_front,0)   #left_front
    gpio.output(left_back,1)    #left_back 
    r_motor.changedutycycle(speed)
    gpio.output(right_front,1)  #right_front
    gpio.output(right_back,0)   #right_back  
def right(speed):
    l_motor.changedutycycle(speed)
    gpio.output(left_front,1)   #left_front
    gpio.output(left_back,0)    #left_back 
    r_motor.changedutycycle(speed)
    gpio.output(right_front,0)  #right_front
    gpio.output(right_back,1)   #right_back
def stop():
    l_motor.changedutycycle(0)
    gpio.output(left_front,0)   #left_front
    gpio.output(left_back,0)    #left_back 
    r_motor.changedutycycle(0)
    gpio.output(right_front,0)  #right_front
    gpio.output(right_back,0)   #right_back
def set_servo_angle(channel,angle):
    angle=4096*((angle*11)+500)/20000
    pwm_servo.set_pwm_freq(50)                #frequency==50hz (servo)
    pwm_servo.set_pwm(channel,0,int(angle)) 
def image_processing():
    # capture the frames
    ret, frame = camera.read()
    # crop the image
    image = frame
    cv2.imshow('frame',frame)
    # to gray
    gray = cv2.cvtcolor(frame, cv2.color_bgr2gray)
    cv2.imshow('gray',gray)
    # gausi blur
    blur = cv2.gaussianblur(gray,(5,5),0)
    #brighten
    blur = cv2.convertscaleabs(blur, none, 1.5, 30)
    #to binary
    ret,binary = cv2.threshold(blur,150,255,cv2.thresh_binary_inv)
    cv2.imshow('binary',binary)
    #close
    kernel = cv2.getstructuringelement(cv2.morph_rect, (17,17))
    close = cv2.morphologyex(binary, cv2.morph_close, kernel)
    cv2.imshow('close',close)
    #get contours
    binary_c,contours,hierarchy = cv2.findcontours(close, 1, cv2.chain_approx_none)
    cv2.drawcontours(image, contours, -1, (255,0,255), 2)
    cv2.imshow('image', image)
    return frame, contours
def get_coord(img, contours):
    image = img.copy()
    try:
        contour = max(contours, key=cv2.contourarea)
        cv2.drawcontours(image, contour, -1, (255,0,255), 2)
        cv2.imshow('new_frame', image)
        # get coord
        m = cv2.moments(contour)
        x = int(m['m10']/m['m00'])
        y = int(m['m01']/m['m00'])
        print(x, y) 
        return x,y        
    except:
        print 'no objects'
        return 0,0    
def move(x,y):
    global second
    #stop
    if x==0 and y==0:
        stop()
    #go ahead
    elif width/4 <x and x<(width-width/4):
        front(70)
    #left
    elif x < width/4:
        left(50)
    #right
    elif x > (width-width/4):
        right(50)   
if __name__ == '__main__':
    init()    
    set_servo_angle(4, 110)     #top servo     lengthwise
    #0:back    180:front    
    set_servo_angle(5, 90)     #bottom servo  crosswise
    #0:left    180:right      
    while 1:
        # 1 image process
        img, contours = image_processing() 
        # 2 get coordinates
        x, y = get_coord(img, contours)
        # 3 move
        move(x,y)       
        # must include this codes(otherwise you can't open camera successfully)
        if cv2.waitkey(1) & 0xff == ord('q'):
            stop()
            gpio.cleanup()    
            break    
    #front(50)
    #back(50)
    #$left(50)
    #right(50)
    #time.sleep(1)
    #stop()
 

检测原理是基于最大轮廓的检测，没有用深度学习的分类，所以容易受到干扰，后期学完深度学习会继续优化。有意见或者想法的朋友欢迎交流。

以上就是opencv物体跟踪树莓派视觉小车实现过程学习的详细内容，更多关于opencv物体跟踪树莓派视觉小车的资料请关注其它相关文章！