Python 如何从边界框中获取图像

如何从该图像中提取边界框？

这是我的密码

tf.disable_v2_behavior()
import cv2
import pathlib

model_path = "/Users/ebayb/Desktop/ndir/Model"

# path to the folder containing images
image_path = "/Users/ebayb/Desktop/top-cube/testing/2"

session = tf.Session(graph=tf.Graph())

tf.saved_model.loader.load(session, ['serve'], model_path)

# extract the coordinates of the rectange that's to be drawn on the image
def draw_boxes(height, width, box, img):
  # starting coordinates of the box
  ymin = int(max(1, (box[0] * height)))
  xmin = int(max(1, (box[1] * width)))
  # last coordinates of the box
  ymax = int(min(height, (box[2] * height)))
  xmax = int(min(width, (box[3] * width)))
  coordinates_list=[ymin, ymax, xmin, xmax]
  print(f"cor: {coordinates_list}")
  # draw a rectange using the coordinates
  cv2.rectangle(img, (xmin, ymin), (xmax, ymax), (10, 255, 0), 2)
  

for file in pathlib.Path(image_path).iterdir():
  # get the current image path
  current_image_path = r"{}".format(file.resolve())
  
  
  img_bytes = open(current_image_path, 'rb').read()
  
  
  result = session.run(['detection_boxes:0', 'detection_scores:0'], feed_dict={'encoded_image_string_tensor:0': [img_bytes]})
  
  boxes = result[0][0]
  scores = result[1][0]
  
  print("For file {}".format(file.stem))

 
  img = cv2.imread(current_image_path)
  
  imH, imW, _ = img.shape
               
  for i in range(len(scores)):
    # only consider a detected object if it's probability is above 25%
    if scores[i] > 0.258:
      print("The box {} has probability {}".format(boxes[i], scores[i]))
      draw_boxes(imH, imW, boxes[i], img)
      
      
      
  
  

  cv2.imwrite("bound.jpg", img)

我想知道如何从中得到边界框？我试过在其他线程上使用其他答案的ROI，但这似乎不起作用，只会得到其他图像

（ROI执行此操作的示例）

以下是ROI的代码：

import numpy as np

# Load image, grayscale, Otsu's threshold 
image = cv2.imread('test.jpg')
original = image.copy()
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
thresh = cv2.threshold(gray, 0, 100, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)[1]

# Find contours, obtain bounding box, extract and save ROI
ROI_number = 0
cnts = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
cnts = cnts[0] if len(cnts) == 2 else cnts[1]
for c in cnts:
    x,y,w,h = cv2.boundingRect(c)
    cv2.rectangle(image, (x, y), (x + w, y + h), (36,255,12), 2)
    ROI = original[y:y+h, x:x+w]
    cv2.imwrite('ROI_{}.png'.format(ROI_number), ROI)
    ROI_number += 1

cv2.imshow('image', image)
cv2.waitKey()

这是我正在使用的测试图像

---

您已经在第一个代码中绘制了矩形。如果需要矩形区域，请修改第一个代码，以使用

xmin、ymin、xmax、ymax

值裁剪原始图像

对于符合

评分[i]>0.258的每个矩形，使用上述值在
绘图框中裁剪原始图像的一部分并保存。那就不需要使用等高线了

对重叠的边界框（如果有）使用
非最大抑制