How to create for loop cv.write in image processing tensorflow and jupyter notebook?-CodePudding

I am currently doing image processing in python and tensorflow want to create a for loop python code for a specified folder This is the code for the specific folder

import pathlib
PATH_TO_TEST_IMAGES_DIR = pathlib.Path('C:/Object_detection/models-master/research/object_detection/test_images')
TEST_IMAGE_PATHS = sorted(list(PATH_TO_TEST_IMAGES_DIR.glob("*.jpg")))
TEST_IMAGE_PATHS

I want to create a for loop in this part of my image processing so that it can output multiple images without right click save as in the jupyter notebook

def show_inference(model, image_path):

  image_np = np.array(Image.open(image_path))

  output_dict = run_inference_for_single_image(model, image_np)

  vis_util.visualize_boxes_and_labels_on_image_array(
      image_np,
      output_dict['detection_boxes'],
      output_dict['detection_classes'],
      output_dict['detection_scores'],
      category_index,
      instance_masks=output_dict.get('detection_masks_reframed', None),
      use_normalized_coordinates=True,
      line_thickness=8)

  display(Image.fromarray(image_np))

CodePudding user response：

Kind of difficult to tell what you're asking for here, but this is a method that would display n images sampled from the directory. The random sampling is done with numpy.random's choice function.

def show_n_inferences(model, image_paths=TEST_IMAGE_PATHS, n=10):
  sampled_paths = np.random.choice(image_paths, size=n)
  for image_path in sampled_paths:
    image_np = np.array(Image.open(image_path))
    output_dict = run_inference_for_single_image(model, image_np)

    vis_util.visualize_boxes_and_labels_on_image_array(
      image_np,
      output_dict['detection_boxes'],
      output_dict['detection_classes'],
      output_dict['detection_scores'],
      category_index,
      instance_masks=output_dict.get('detection_masks_reframed', None),
      use_normalized_coordinates=True,
      line_thickness=8)

    display(Image.fromarray(image_np))