python中在用mediapipe-pose部分的时候，遇到一个问题，如何解决?

import cv2
import mediapipe as me
#调用摄像头
cap = cv2.VideoCapture(0,cv2.CAP_DSHOW)
#姿态识别
pose_recognition =me.solutions.pose
#检测摄像头
while (cap.isOpened()):
    #获取摄像头每帧图像
    ret,frame = cap.read()
    frame= cv2.flip(frame,1)
    #转换读取图像的格式(rgb)
    imgs=cv2.cvtColor(frame,cv2.COLOR_BGR2RGB)
    result =pose_recognition.Pose.process(imgs)
    if results.pose_landmarks:
        mp.solutions.drawing_utils.draw_landmarks(imgs, results.pose_landmarks,pose_recognition.POSE_CONNECTIONS)
        for index, lm in enumerate(results.pose_landmarks.landmark):
            h, w, c = imgs.shape

    #显示所获取的摄像头的每帧图像
    cv2.imshow("pose shibie",frame)
    k = cv2.waitKey(1)
    #按下空格后进入
    if k==32:
        #关闭摄像头
        cap.release()
        break
-----------------------------------------------------------------------------------------------------------------------------------
Traceback (most recent call last):
  File "D:\pythonProject\venv\share\pose.py", line 14, in <module>
    result =pose_recognition.Pose.process(imgs)
TypeError: process() missing 1 required positional argument: 'image'

10行下面判断下ret是否为True，你这个意思就是图片为空