来源网址：https://blog.csdn.net/GuoZhen_Zhou/article/details/114836660

import cv2
import numpy as np
# 乒乓球位置识别，加入了指示移动方向的箭头


def empty(a):
    pass


def draw_direction(img, lx, ly, nx, ny):
    # 根据上一位置与当前位置计算移动方向并绘制箭头
    dx = nx - lx
    dy = ny - ly
    if abs(dx) < 4 and abs(dy) < 4:
        dx = 0
        dy = 0
    else:
        r = (dx**2 + dy**2)**0.5
        dx = int(dx/r*40)
        dy = int(dy/r*40)
        # print(dx, dy)
    cv2.arrowedLine(img, (60, 100), (60+dx, 100+dy), (0, 255, 0), 2)
    # print(nx-lx, ny-ly)   # 噪声一般为+-1
    # cv2.arrowedLine(img, (150, 150), (150+(nx-lx)*4, 150+(ny-ly)*4), (0, 0, 255), 2, 0, 0, 0.2)


frameWidth = 640
frameHeight = 480
cap = cv2.VideoCapture(0)  # 0对应笔记本自带摄像头
cap.set(3, frameWidth)  # set中，这里的3，下面的4和10是类似于功能号的东西，数字的值没有实际意义
cap.set(4, frameHeight)
cap.set(10, 80)        # 设置亮度
pulse_ms = 30

# 调试用代码，用来产生控制滑条
# cv2.namedWindow("HSV")
# cv2.resizeWindow("HSV", 640, 300)
# cv2.createTrackbar("HUE Min", "HSV", 4, 179, empty)
# cv2.createTrackbar("SAT Min", "HSV", 180, 255, empty)
# cv2.createTrackbar("VALUE Min", "HSV", 156, 255, empty)
# cv2.createTrackbar("HUE Max", "HSV", 32, 179, empty)
# cv2.createTrackbar("SAT Max", "HSV", 255, 255, empty)
# cv2.createTrackbar("VALUE Max", "HSV", 255, 255, empty)

lower = np.array([18, 180, 156])     # 适用于橙色乒乓球4<=h<=32
upper = np.array([43, 255, 255])
# lower = np.array([172, 113, 118])     # 适用于红色乒乓球
# upper = np.array([179, 255, 255])

targetPos_x = 0
targetPos_y = 0
lastPos_x = 0
lastPos_y = 0

while True:
    _, img = cap.read()

    imgHsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)

    # h_min = cv2.getTrackbarPos("HUE Min", "HSV")
    # h_max = cv2.getTrackbarPos("HUE Max", "HSV")
    # s_min = cv2.getTrackbarPos("SAT Min", "HSV")
    # s_max = cv2.getTrackbarPos("SAT Max", "HSV")
    # v_min = cv2.getTrackbarPos("VALUE Min", "HSV")
    # v_max = cv2.getTrackbarPos("VALUE Max", "HSV")
    #
    # lower = np.array([h_min, s_min, v_min])
    # upper = np.array([h_max, s_max, v_max])

    imgMask = cv2.inRange(imgHsv, lower, upper)     # 获取遮罩
    imgOutput = cv2.bitwise_and(img, img, mask=imgMask)
    contours, hierarchy = cv2.findContours(imgMask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)   # 查找轮廓
    # https://blog.csdn.net/laobai1015/article/details/76400725
    # CV_RETR_EXTERNAL 只检测最外围轮廓
    # CV_CHAIN_APPROX_NONE 保存物体边界上所有连续的轮廓点到contours向量内
    imgMask = cv2.cvtColor(imgMask, cv2.COLOR_GRAY2BGR)     # 转换后，后期才能够与原画面拼接，否则与原图维数不同

    # 下面的代码查找包围框，并绘制
    x, y, w, h = 0, 0, 0, 0
    for cnt in contours:
        area = cv2.contourArea(cnt)
        # print(area)
        if area > 300:
            x, y, w, h = cv2.boundingRect(cnt)
            lastPos_x = targetPos_x
            lastPos_y = targetPos_y
            targetPos_x = int(x+w/2)
            targetPos_y = int(y+h/2)
            cv2.rectangle(img, (x, y), (x+w, y+h), (0, 255, 0), 2)
            cv2.circle(img, (targetPos_x, targetPos_y), 2, (0, 255, 0), 4)

    # 坐标（图像内的）
    cv2.putText(img, "({:0<2d}, {:0<2d})".format(targetPos_x, targetPos_y), (20, 30),
                cv2.FONT_HERSHEY_PLAIN, 1, (0, 255, 0), 2)  # 文字
    draw_direction(img, lastPos_x, lastPos_y, targetPos_x, targetPos_y)

    imgStack = np.hstack([img, imgOutput])
    # imgStack = np.hstack([img, imgMask])            # 拼接
    cv2.imshow('Horizontal Stacking', imgStack)     # 显示
    if cv2.waitKey(pulse_ms) & 0xFF == ord('q'):          # 按下“q”推出（英文输入法）
        print("Quit\n")
        break

cap.release()
cv2.destroyAllWindows()