在 Tensorflow 对象检测 API 中裁剪图像并显示它

Question

我正在使用 tensorflow 异议检测 API 来检测护照上的 MRZ 代码。我已经对数据进行了训练，一切正常。它用一个边界框完美地识别了它周围的机读区代码。但是，现在我只想裁剪 MRZ 代码（或边界框），但在使用 PIL 图像库时遇到了麻烦。这就是我的代码的样子：

# Import packages
import os
import cv2
import numpy as np
import tensorflow as tf
import sys
from PIL import Image

# This is needed since the notebook is stored in the object_detection folder.
sys.path.append("..")

# Import utilites
from utils import label_map_util
from utils import visualization_utils as vis_util

# Name of the directory containing the object detection module we're using
MODEL_NAME = 'inference_graph'
IMAGE_NAME = 'test.jpg'

# Grab path to current working directory
CWD_PATH = os.getcwd()

# Path to frozen detection graph .pb file, which contains the model that is used
# for object detection.
PATH_TO_CKPT = os.path.join(CWD_PATH,MODEL_NAME,'frozen_inference_graph.pb')

# Path to label map file
PATH_TO_LABELS = os.path.join(CWD_PATH,'training','labelmap.pbtxt')

# Path to image
PATH_TO_IMAGE = os.path.join(CWD_PATH,IMAGE_NAME)

# Number of classes the object detector can identify
NUM_CLASSES = 6

# Load the label map.
# Label maps map indices to category names, so that when our convolution
# network predicts `5`, we know that this corresponds to `king`.
# Here we use internal utility functions, but anything that returns a
# dictionary mapping integers to appropriate string labels would be fine
label_map = label_map_util.load_labelmap(PATH_TO_LABELS)
categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)
category_index = label_map_util.create_category_index(categories)

# Load the Tensorflow model into memory.
detection_graph = tf.Graph()
with detection_graph.as_default():
    od_graph_def = tf.GraphDef()
    with tf.gfile.GFile(PATH_TO_CKPT, 'rb') as fid:
        serialized_graph = fid.read()
        od_graph_def.ParseFromString(serialized_graph)
        tf.import_graph_def(od_graph_def, name='')

    sess = tf.Session(graph=detection_graph)

# Define input and output tensors (i.e. data) for the object detection classifier

# Input tensor is the image
image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')

# Output tensors are the detection boxes, scores, and classes
# Each box represents a part of the image where a particular object was detected
detection_boxes = detection_graph.get_tensor_by_name('detection_boxes:0')

# Each score represents level of confidence for each of the objects.
# The score is shown on the result image, together with the class label.
detection_scores = detection_graph.get_tensor_by_name('detection_scores:0')
detection_classes = detection_graph.get_tensor_by_name('detection_classes:0')

# Number of objects detected
num_detections = detection_graph.get_tensor_by_name('num_detections:0')

# Load image using OpenCV and
# expand image dimensions to have shape: [1, None, None, 3]
# i.e. a single-column array, where each item in the column has the pixel RGB value
image = cv2.imread(PATH_TO_IMAGE)
image_expanded = np.expand_dims(image, axis=0)

# Perform the actual detection by running the model with the image as input
(boxes, scores, classes, num) = sess.run(
    [detection_boxes, detection_scores, detection_classes, num_detections],
    feed_dict={image_tensor: image_expanded})


# Draw the results of the detection (aka 'visulaize the results')

vis_util.visualize_boxes_and_labels_on_image_array(
    image,
    np.squeeze(boxes),
    np.squeeze(classes).astype(np.int32),
    np.squeeze(scores),
    category_index,
    use_normalized_coordinates=True,
    line_thickness=8,
    min_score_thresh=0.80)



width, height = image.shape[:2]
for i, box in enumerate(np.squeeze(boxes)):
      if(np.squeeze(scores)[i] > 0.80):
        (ymin, xmin, ymax, xmax) = (box[0]*height, box[1]*width, box[2]*height, box[3]*width)
        im = Image.open('test.jpg')
        im.crop((ymin, xmin, ymax - ymax, xmax - xmin)).show()

# All the results have been drawn on image. Now display the image.
cv2.imshow('Object detector', image)

# Press any key to close the image
cv2.waitKey(0)

# Clean up
cv2.destroyAllWindows()


However, I always get the following error:

Traceback (most recent call last):
  File "/usr/local/lib/python3.6/site-packages/PIL/ImageFile.py", line 482, in _save
    fh = fp.fileno()
AttributeError: '_idat' object has no attribute 'fileno'

在处理上述异常的过程中，又发生了一个异常：

Traceback (most recent call last):
  File "object_detection_image.py", line 128, in <module>
    im.crop((ymin, xmin, ymax - ymax, xmax - xmin)).show()
  File "/usr/local/lib/python3.6/site-packages/PIL/Image.py", line 1977, in show
    _show(self, title=title, command=command)
  File "/usr/local/lib/python3.6/site-packages/PIL/Image.py", line 2802, in _show
    _showxv(image, **options)
  File "/usr/local/lib/python3.6/site-packages/PIL/Image.py", line 2807, in _showxv
    ImageShow.show(image, title, **options)
  File "/usr/local/lib/python3.6/site-packages/PIL/ImageShow.py", line 51, in show
    if viewer.show(image, title=title, **options):
  File "/usr/local/lib/python3.6/site-packages/PIL/ImageShow.py", line 75, in show
    return self.show_image(image, **options)
  File "/usr/local/lib/python3.6/site-packages/PIL/ImageShow.py", line 95, in show_image
    return self.show_file(self.save_image(image), **options)
  File "/usr/local/lib/python3.6/site-packages/PIL/ImageShow.py", line 91, in save_image
    return image._dump(format=self.get_format(image), **self.options)
  File "/usr/local/lib/python3.6/site-packages/PIL/Image.py", line 624, in _dump
    self.save(filename, format, **options)
  File "/usr/local/lib/python3.6/site-packages/PIL/Image.py", line 1930, in save
    save_handler(self, fp, filename)
  File "/usr/local/lib/python3.6/site-packages/PIL/PngImagePlugin.py", line 821, in _save
    [("zip", (0, 0)+im.size, 0, rawmode)])
  File "/usr/local/lib/python3.6/site-packages/PIL/ImageFile.py", line 490, in _save
    e.setimage(im.im, b)
SystemError: tile cannot extend outside image

我已尝试多次更改数字。有时裁剪后的图片只是黑色。尝试搜索谷歌，但似乎无法获得边界框。非常感谢您的帮助。

Answer 1

@Rochan 几乎完美地回答了这个问题。我想指出您代码中的一个错误。

      ymin  = boxi[0]
      xmin  = boxi[1]
      ymax  = boxi[2]
      xmax  = boxi[3]

（这是正确的格式）

非常感谢@Rochan

Answer 2

你需要做的就是在指定路径models/research/object detection/utils/visualizatioin_utils.py中找到文件

做以下事情

import cv2

count = 0

找到一个函数

      def draw_bounding_box_on_image():
             draw = ImageDraw.Draw(image):
             arr = np.array(image)
             im_width, im_height = image.size

 if use_normalized_coordinates:
    (left, right, top, bottom) = (xmin * im_width, xmax * im_width,
                              ymin * im_height, ymax * im_height)
     a,b,c,d = int(left) , int(right) , int(top) ,int(bottom)
     arr = arr[c:d,a:b]
     global count
     cv2.imwrite("yourpath/QRCODE{}.jpg".format(count),arr)
     count = count+1

 else:
    (left, right, top, bottom) = (xmin, xmax, ymin, ymax)
    draw.line([(left, top), (left, bottom), (right, bottom),
         (right, top), (left, top)], width=thickness, fill=color

Answer 3

我在visualization_utils.py的最后添加了如下代码，

找到函数“visualize_boxes_and_labels_on_image_array” 然后我将检测到的框附加到一个数组中。

for i in range(min(max_boxes_to_draw, boxes.shape[0])):
   if scores is None or scores[i] > min_score_thresh:
      box = tuple(boxes[i].tolist())
      bigger_bounding_box.append(box)

我添加了以下代码：

bigger_bounding_box=[]
count =0
for i in range(0,len(bigger_bounding_box)):
        im_width, im_height = image.shape[0], image.shape[1]
        arr = numpy.array(image)
        box = bigger_bounding_box[i]
        xmin  = box[0]
        ymin  = box[1]
        xmax  = box[2]
        ymax  = box[3]
        (left, right, top, bottom) = (xmin * im_width, xmax * im_width, ymin * im_height, ymax * im_height)
        a,b,c,d = int(left) , int(right) , int(top) ,int(bottom)
        arr = arr[c:d,a:b]
        cv2.imwrite("C:/Users/tensor19/Desktop/images/QRCODE{}.jpg".format(count),arr)
        count =count+1

模型能够检测到输入图像中的对象，但它没有裁剪在输入图像中检测到的相同对象。