这个函数前两个我知道是图片相移的横纵方向的数据，但是这第三个数据（个人认为应该是角度）该怎么转换成弧度制等常用的单位啊。个人尝试但好像效果不太好。。用过的大神能指点一下吗orz 整个代码如下，主要是检测条纹图形的相移（位移加旋转）

# -*-coding:utf-8 -*-  
__author__ = "ZJL"
 
import cv2
import time
import numpy as np
 
# 保存截图
save_path = './img/'
 
# 定义摄像头对象，其参数0表示第一个摄像头
camera = cv2.VideoCapture("moving_updown.mp4")
 
# 判断视频是否打开
if (camera.isOpened()):
    print('Open')
else:
    print('摄像头未打开')
 
# 测试用,查看视频size
size = (int(camera.get(cv2.CAP_PROP_FRAME_WIDTH)),
        int(camera.get(cv2.CAP_PROP_FRAME_HEIGHT)))
print('size:'+repr(size))
 
# 帧率
fps = 60
# 总是取前一帧做为背景（不用考虑环境影响）
pre_frame = None

#计时，绘制图表
cnt = 0
dt = 2
dot = np.array([[1400,400],[0,400]])
img = np.zeros((800,1600,3), np.uint8)
pre_ddyy = 0
pre_ddrr = 0

cv2.polylines(img,[dot],True,(0,0,153))

while(1):
    start = time.time()
    cnt = cnt+dt
    
    # 读取视频流
    ret, frame = camera.read()

    if not ret:
        break
    end = time.time()
    
    # 转灰度图
    gray_lwpCV = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
 
    # 显示图像
    cv2.imshow("capture", frame)
 
    # 运动检测部分
    seconds = end - start
    if seconds < 1.0 / fps:
        time.sleep(1.0 / fps - seconds)
    gray_lwpCV = cv2.resize(gray_lwpCV, (500, 500))
    # 用高斯滤波进行模糊处理
    gray_lwpCV = cv2.GaussianBlur(gray_lwpCV, (3, 3), 10)
 
    # 如果没有背景图像就将当前帧当作背景图片
    if pre_frame is None:
        pre_frame = gray_lwpCV
    else:
        # absdiff把两幅图的差的绝对值输出到另一幅图上面来
        img_delta = cv2.absdiff(pre_frame, gray_lwpCV)
        #threshold阈值函数(原图像应该是灰度图,对像素值进行分类的阈值,当像素值高于（有时是小于）阈值时应该被赋予的新的像素值,阈值方法)
        thresh = cv2.threshold(img_delta, 25, 255, cv2.THRESH_BINARY)[1]
        # 膨胀图像
        thresh = cv2.dilate(thresh, None, iterations=2)
        # findContours检测物体轮廓(寻找轮廓的图像,轮廓的检索模式,轮廓的近似办法)
        contours, hierarchy = cv2.findContours(thresh.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

        ddxy = [0.0,0.0]
        arrow = np.zeros((300,300,3),np.uint8)
        
        for c in contours:
            # 设置敏感度
            # contourArea计算轮廓面积
            if cv2.contourArea(c) < 0:
                continue
            else:
                # 画出矩形框架,返回值x，y是矩阵左上点的坐标，w，h是矩阵的宽和高
                (x, y, w, h) = cv2.boundingRect(c)
                # rectangle(原图,(x,y)是矩阵的左上点坐标,(x+w,y+h)是矩阵的右下点坐标,(0,255,0)是画线对应的rgb颜色,2是所画的线的宽度)
                cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
                # putText 图片中加入文字
                cv2.putText(frame, "now time: {}".format(str(time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time()))) ), (10, 20),
                            cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 2)
                print("出现目标物，请求核实")



            #就这里，wtf存储的是疑似是角度的数据  
                print("now time: {}".format(str(time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time()))) ))
                ddxy, wtf = cv2.phaseCorrelate(np.float32(pre_frame), np.float32(gray_lwpCV))
                print(wtf)
                cv2.arrowedLine(frame,(200,200),(200+int(ddxy[0]*20),200+int(ddxy[1]*20)), (255, 0, 0), thickness=2, line_type=0, shift=0, tipLength=0.3)
                dot = np.insert(dot,-1,values = [cnt+dt,100+int(ddxy[1]*20)],axis = 0)
                
                
                cv2.polylines(img,[np.array([[cnt,pre_ddyy],[cnt+dt,400+int(ddxy[1]*20)]])],True,(0,0,255))
                cv2.polylines(img,[np.array([[cnt,pre_ddrr],[cnt+dt,400+int(wtf*180)]])],True,(23,123,255))
                
                cv2.imshow('line',img)
                
                pre_ddyy = 400+int(ddxy[1]*20)
                pre_ddrr = 400+int(wtf*180)


                
                # 保存图像
                cv2.imwrite(save_path + str(time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time()))) + '.jpg', frame)
                break
        pre_frame = gray_lwpCV
 
        # 显示图像
     
        cv2.imshow("capture", frame)
        #cv2.imshow("capture", arrow)
        
        # cv2.imshow("Thresh", thresh)
        # 进行阀值化来显示图片中像素强度值有显著变化的区域的画面
        cv2.imshow("Frame Delta", img_delta)
 
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

 
#destroyAllWindows()关闭所有图像窗口
#cv2.destroyAllWindows()