Execute model with segemtation

[Petopp]

Hello,
does anyone have an example of how to run a seg model to recognise an image from a webcam.

I've got that far, but that's as far as I can go:

import cv2
import yolov5

Pfad_Modell=r"C:\Users\toppel\...\Modelle\Yolo_v5\Seg_Test\exp\weights\best.pt"
model = yolov5.load(Pfad_Modell)


cap = cv2.VideoCapture(0)
cap.set (cv2.CAP_PROP_BUFFERSIZE,0)
cap.set(cv2.CAP_PROP_POS_FRAMES, 1)

while (True):
      cap.set(cv2.CAP_PROP_POS_FRAMES, 1)
      ret, img = cap.read()
      
      results = model(img)
      
            
      cv2.imshow("Ausgabe",img)
      

      if cv2.waitKey(1) & 0xFF == ord('q'):
            break

I can see that the system is also doing something via CUDA, but I can't get the results. Do any of you know what to do?