Overhaul to use stochastic pipeline method.

Augmentor/ImageSource.py

@@ -1,165 +1,29 @@
-from __future__ import print_function
-from __future__ import unicode_literals
-from __future__ import division
-from __future__ import absolute_import
-from builtins import next
-from builtins import str
-from future import standard_library
-standard_library.install_aliases()
+from __future__ import (absolute_import, division,
+                        print_function, unicode_literals)
+from builtins import *
 
-# Module imports
-from . import ImageDetails
-from . import ImageFormat
-from . import ProgramFinishedException
-
-from PIL import Image
-from terminaltables import GithubFlavoredMarkdownTable
-from fractions import gcd
 import os
-import glob2
-
-class ImageSource(object):
-    def __init__(self, root_path):
-        self.root_path = root_path
-        self.list_of_images = []
-        self.image_source_slice = None
-        self.processed_images = 0
-        self.file_extension = ('jpg', 'jpeg', 'png', 'bmp')
-        self.dimensions = {}
-        self.image_format = ImageFormat.ImageFormat()
-        self.scan(root_path)
-        self.image_iterator = iter(self.list_of_images)
-        if len(self.list_of_images) == 0:
-            raise ProgramFinishedException.ProgramFinishedException("Image repository is empty.")
-
-    def scan(self, path_to_scan):
-        # for root, subdirs, list_of_files in os.walk(path_to_scan):
-        list_of_files = []
-
-        if isinstance(path_to_scan, list):
-            print("Processing list of images")
-            # List of PIL images
-            if all(isinstance(item, Image.Image) for item in path_to_scan):
-                for file in path_to_scan:
-                    print(file.filename)
-                    self.list_of_images.append(ImageDetails.ImageDetails(file, file.filename))
-                self.root_path = os.path.dirname(path_to_scan[0].filename)
-
-            # List of images paths
-            elif all(os.path.isfile(item) for item in path_to_scan):
-                for file in path_to_scan:
-                    image = Image.open(file)
-                    self.list_of_images.append(ImageDetails.ImageDetails(image, file))
-                    del image
-                self.root_path = os.path.dirname(path_to_scan[0])
-            else:
-                raise ProgramFinishedException.ProgramFinishedException(
-                    "List contains unsupported formats. Only list of PIL-images or list of direct paths supported!")
-
-        elif isinstance(path_to_scan, Image.Image):
-            print("Processing PIL image object")
-            self.root_path = os.path.dirname(path_to_scan.filename)
-            self.list_of_images.append(ImageDetails.ImageDetails(path_to_scan, path_to_scan.filename))
-
-        elif any(extension in path_to_scan for extension in self.file_extension):
-            print("Processing one image")
-            self.root_path = os.path.dirname(path_to_scan)
-            image = Image.open(path_to_scan)
-            self.list_of_images.append(ImageDetails.ImageDetails(image, path_to_scan))
-            del image
-
-        elif os.path.isdir(path_to_scan):
-            print("Processing directory")
-            for extension in self.file_extension:
-                list_of_files.extend(glob2.glob(path_to_scan + '/**/*.' + extension))
-            for file in list_of_files:
-                image = Image.open(file)
-                self.list_of_images.append(ImageDetails.ImageDetails(image, file))
-                del image
-        else:
-            raise ProgramFinishedException.ProgramFinishedException(
-                "Unsupported image source. Please have look at the documentation!")
-
-    def populate_images(self):
-        for image in self.list_of_images:
-            image.populateImage()
-        return 0
+import glob
 
-    def populate_one_image(self):
-        try:
-            image = next(self.image_iterator)
-            image.populateImage()
-            return image
-        except StopIteration:
-            return -1
 
-    def setup_summary(self):
-        for image in self.list_of_images:
-
-            if image.dimensions in self.dimensions:
-                self.dimensions[image.dimensions] += 1
-            else:
-                self.dimensions[image.dimensions] = 1
-
-    def summary(self):
-        self.setup_summary()
-        file_types = self.process_filetypes()
-        table_data = [
-            ['ImageSource Summary:', ''],
-            ['Image count', str(len(self.list_of_images))],
-            ['Dimensions', self.process_dimensions()],
-            ['Aspect ratio(s)', self.process_aspect_ratio()],
-            ['File type(s)', file_types[1]],
-            ['File extension(s)', file_types[0]],
-            ['File format(s)', self.process_file_formats()[0]],
-        ]
-        table = GithubFlavoredMarkdownTable(table_data)
-        print(table.table)
-
-    def process_dimensions(self):
-        result = ""
-        for dimension in sorted(self.dimensions)[:2]:
-            result += str(dimension[0]) + 'x' + str(dimension[1]) + '(' + str(self.dimensions.get(dimension)) + ')\n'
-
-        if len(self.dimensions) > 2:
-            result += 'and more...\n'
-
-        return result[:-1]
-
-    def process_file_formats(self):
-        file_formats = {}
-        for image in self.list_of_images:
-            if image.mode not in file_formats:
-                file_formats[image.mode] = self.image_format.get_format_printable(image.mode)
-
-        return [str(', '.join(str(file_format) for file_format in file_formats.values())), str(len(file_formats))]
-
-    def process_filetypes(self):
-        file_types = []
-        for image in self.list_of_images:
-            if image.extension not in file_types:
-                file_types.append(image.extension)
+class ImageSource(object):
+    """
+    The ImageSource class is used to search for and contain paths to images for augmentation.
+    """
+    def __init__(self, source_directory, recursive_scan=False):
+        source_directory = os.path.abspath(source_directory)
+        self.image_list = self.scan_directory(source_directory, recursive_scan)
 
-        return [str(', '.join(str(file_type[1:]) for file_type in file_types)), str(len(file_types))]
+        self.largest_file_dimensions = (800, 600)
 
-    def process_aspect_ratio(self):
-        aspect_ratios = []
-        for width, height in self.dimensions.keys():
-            ratio = gcd(width, height)
-            if [width // ratio, height // ratio] not in aspect_ratios:
-                aspect_ratios.append([width // ratio, height // ratio])
+    def scan_directory(self, source_directory, recusrive_scan=False):
+        # TODO: Make this a static member somewhere later
+        file_types = ['*.jpg', '*.bmp', '*.jpeg', '*.gif', '*.img', '*.png']
+        file_types.extend([str.upper(x) for x in file_types])
 
-        result = '\n'.join(str(ratio[0]) + ':' + str(ratio[1]) for ratio in aspect_ratios[:2])
-        if len(aspect_ratios) > 2:
-            result += '\nand more...'
+        list_of_files = []
 
-        return result
+        for file_type in file_types:
+            list_of_files.extend(glob.glob(os.path.join(os.path.abspath(source_directory), file_type)))
 
-    def process_list(self, list_of_images):
-        if all(isinstance(item, Image.Image) for item in list_of_images):
-            return
-        elif all(os.path.isfile(item) for item in list_of_images):
-            return
-        else:
-            raise ProgramFinishedException.ProgramFinishedException(
-                "List contains unsupported formats. Only list of PIL-images or list of direct paths supported!")
+        return list_of_files

Augmentor/ImageUtilities.py

@@ -0,0 +1,30 @@
+from __future__ import (absolute_import, division,
+                        print_function, unicode_literals)
+from builtins import *
+
+import os
+import glob
+
+
+class AugmentorImage(object):
+    def __init__(self, image_path):
+        self.image_path = image_path
+
+
+def extract_paths_and_extensions(image_path):
+    file_name, extension = os.path.splitext(image_path)
+    root_path = os.path.dirname(image_path)
+
+    return file_name, extension, root_path
+
+
+def scan_directory(source_directory, recursive_scan=False):
+    file_types = ['*.jpg', '*.bmp', '*.jpeg', '*.gif', '*.img', '*.png']
+    file_types.extend([str.upper(x) for x in file_types])
+
+    list_of_files = []
+
+    for file_type in file_types:
+        list_of_files.extend(glob.glob(os.path.join(os.path.abspath(source_directory), file_type)))
+
+    return list_of_files