在程序中获取image的断言失败错误

Question

我创建了一个api来托管这个程序，但是在托管平台上得到断言失败错误，而程序在我的本地主机上正常运行。

import cv2
import mediapipe as mp
import numpy as np
from flask import Flask, render_template, Response

app=Flask(__name__)
def gen():
mpDraw = mp.solutions.drawing_utils
mp_pose = mp.solutions.pose
pose = mp_pose.Pose()

cap = cv2.VideoCapture(0)
counter=0
stage=None
while True:
success, img = cap.read()
imgRGB = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
result = pose.process(imgRGB)

if result.pose_landmarks:
mpDraw.draw_landmarks(img, result.pose_landmarks, mp_pose.POSE_CONNECTIONS)
landmarks = result.pose_landmarks.landmark

shoulder = [landmarks[mp_pose.PoseLandmark.LEFT_SHOULDER.value].x, landmarks[mp_pose.PoseLandmark.LEFT_SHOULDER.value].y]
elbow = [landmarks[mp_pose.PoseLandmark.LEFT_ELBOW.value].x, landmarks[mp_pose.PoseLandmark.LEFT_ELBOW.value].y]
wrist = [landmarks[mp_pose.PoseLandmark.LEFT_WRIST.value].x, landmarks[mp_pose.PoseLandmark.LEFT_ELBOW.value].y]
hip = [landmarks[mp_pose.PoseLandmark.LEFT_HIP.value].x, landmarks[mp_pose.PoseLandmark.LEFT_HIP.value].y]

a = np.array(shoulder) 
b = np.array(elbow) 
c = np.array(wrist) 
d = np.array(hip)
ang = 0
ang1 = 0
radians= np.arctan2(c[1]-b[1], c[0]-b[0]) - np.arctan2(a[1]-b[1], a[0]-b[0])
ang = np.abs(radians*180/np.pi)
if ang >180.0:
ang = 360-ang

radians1= np.arctan2(b[1]-a[1], b[0]-a[0]) - np.arctan2(d[1]-a[1], d[0]-a[0])
ang1 = np.abs(radians1*180/np.pi)
if ang1 >180.0:
ang1 = 360-ang1


cv2.putText(img, str(ang), 
tuple(np.multiply(elbow, [640, 480]).astype(int)), 
cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 255, 255), 2, cv2.LINE_AA
)
cv2.putText(img, str(ang1), 
tuple(np.multiply(shoulder, [640, 480]).astype(int)), 
cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 255, 255), 2, cv2.LINE_AA
)






if ang1<50 and 109<ang:
stage="DOWN"

if ang1<50 and ang<70 and stage=='DOWN':
stage="UP"
counter=counter+1


else:    
mpDraw.draw_landmarks(img, result.pose_landmarks,
mp_pose.POSE_CONNECTIONS,
mpDraw.DrawingSpec(color=(245,117,66), thickness=2, circle_radius=2),
mpDraw.DrawingSpec(color=(245,66,230), thickness=2, circle_radius=2))


cv2.putText(img, 'REPS', (15,12), 
cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0,0,0), 1, cv2.LINE_AA)
cv2.putText(img, str(counter), 
(10,60), 
cv2.FONT_HERSHEY_SIMPLEX, 2, (255,255,255), 2, cv2.LINE_AA)

cv2.putText(img, 'STAGE', (65,12), 
cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0,0,0), 1, cv2.LINE_AA)
cv2.putText(img, stage, 
(60,60), 
cv2.FONT_HERSHEY_SIMPLEX, 2, (255,255,255), 2, cv2.LINE_AA)


#cv2.imshow("Pose detection", img)
frame = cv2.imencode('.jpg', img)[1].tobytes()
yield (b'--framern'b'Content-Type: image/jpegrnrn' + frame + b'rn')
if cv2.waitKey(20) & 0xFF == ord('q') :
break




@app.route('/',methods=['GET'])
def video_feed():
return Response(gen(),mimetype ='multipart/x-mixed-replace; boundary=frame')

if __name__=="__main__":
app.run(debug=False)

运行WSGI应用程序出错cv2。_src.empty() in function 'cvtColor'

File "/usr/local/lib/python3.9/site-packages/werkzeug/wsgi.py"，第462行，在下一个返回self._next ()文件"/usr/local/lib/python3.9/site-packages/werkzeug/wrappers/response.py"，第49行，在_iter_encoded对于iterable中的item:

File "/home/shivam1673/mysite/flask_app.py"，第20行，genImg = cv2。cv2.COLOR_BGR2RGB cvtColor(图片)

Answer 1

我在我的机器上运行了这段代码，它运行得很好。此错误通常意味着您将一个空Numpy数组作为函数的参数。在你的例子中，你把img给了cv2.cvtColor。

我建议检查两件事:

1. 检查您的VideoCapture是否正在工作-您可以使用调试器执行此操作，例如，查看是否一切正常。
2. 注意视频捕获的success输出。如果将success设置为False，则意味着没有返回图像，因此img将不包含正确的图像。您没有包含这些信息，这可能是您出错的原因。

一个简单的检查方法是:

...
while True:
success, img = cap.read()
if not success:
continue # handle the error case
...