yolo8 masks detection gives me multiple contours

Question

I edited the question after i experimented a bit more !

I am experimenting with yolo8.

I made a test script to show the problem:

#
#  test script to explore multi contours for 1 found object.
import numpy as np
import cv2

img = cv2.imread('demo.jpg')
imgcopy = img.copy()


from ultralytics import YOLO

# Load a model
model = YOLO('yolov8n-seg.pt')  # load an official model
#model = YOLO('path/to/best.pt')  # load a custom model

# Predict with the model
results = model(img, retina_masks=True, save=True, imgsz = 640, conf=0.5, boxes=False, show_labels=False, show_conf=False, save_crop=True, save_txt=True, save_conf=True)  # predict on an image



for result in results:
    masks = result.masks
    for mask in masks.data:
        mask_np= mask.cpu().numpy().astype(np.uint8)
        #contours, _  = cv2.findContours(mask_np, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
        contours, _  = cv2.findContours(mask_np, cv2.RETR_TREE, cv2.CHAIN_APPROX_NONE)

        print ('We found ' + str(len(contours)) + ' objects')

        # i know i get 3 contours so this works !
        imgcopy = cv2.drawContours(imgcopy, contours=contours[0], contourIdx=-1, color=(0, 255, 0), thickness=2, lineType=cv2.LINE_AA)
        imgcopy = cv2.drawContours(imgcopy, contours=contours[1], contourIdx=-1, color=(255, 255, 0), thickness=2, lineType=cv2.LINE_AA)
        imgcopy = cv2.drawContours(imgcopy, contours=contours[2], contourIdx=-1, color=(255,0, 255), thickness=2, lineType=cv2.LINE_AA)

cv2.imwrite('result.jpg', imgcopy)

I expected to find 1 contour but i find 3 contours in this example.

input image:

output image:

So 1 mask belongs to 1 found object but it has 3 contours.

Why is it finding multiple contours ?

How should i filter out the correct contour ? I found that the first found contour is not allways the correct one.