andrewdcampbell · gregorylearns · Oct 12, 2022 · Oct 12, 2022 · Oct 12, 2022 · Oct 12, 2022
diff --git a/README.md b/README.md
@@ -9,19 +9,19 @@ On my test dataset of 280 images, the program correctly detected the corners of
 This project makes use of the transform and imutils modules from pyimagesearch (which can be accessed [here](http://www.pyimagesearch.com/2014/09/01/build-kick-ass-mobile-document-scanner-just-5-minutes/)). The UI code for the interactive mode is adapted from `poly_editor.py` from [here](https://matplotlib.org/examples/event_handling/poly_editor.html).
 
 * You can manually click and drag the corners of the document to be perspective transformed:
-![Example of interactive GUI](https://github.com/andrewdcampbell/doc_scanner/blob/master/ui.gif)
+![Example of interactive GUI](readme_media/ui.gif)
 
 * The scanner can also process an entire directory of images automatically and save the output in an output directory:
-![Image Directory of images to be processed](https://github.com/andrewdcampbell/doc_scanner/blob/master/before_after.gif)
+![Image Directory of images to be processed](readme_media/before_after.gif)
 
 #### Here are some examples of images before and after scan:
-<img src="https://github.com/andrewdcampbell/doc_scanner/blob/master/sample_images/cell_pic.jpg" height="450"> <img src="https://github.com/andrewdcampbell/doc_scanner/blob/master/output/cell_pic.jpg" height="450">
+<img src="sample_images/cell_pic.jpg" height="450"> <img src="sample_output/cell_pic.jpg" height="450">
 
-<img src="https://github.com/andrewdcampbell/doc_scanner/blob/master/sample_images/receipt.jpg" height="450"> <img src="https://github.com/andrewdcampbell/doc_scanner/blob/master/output/receipt.jpg" height="450">
+<img src="sample_images/receipt.jpg" height="450"> <img src="sample_output/receipt.jpg" height="450">
 
-<img src="https://github.com/andrewdcampbell/doc_scanner/blob/master/sample_images/math_cheat_sheet.JPG" height="450"> <img src="https://github.com/andrewdcampbell/doc_scanner/blob/master/output/math_cheat_sheet.JPG" height="450">
+<img src="sample_images/math_cheat_sheet.JPG" height="450"> <img src="sample_output/math_cheat_sheet.JPG" height="450">
 
-<img src="https://github.com/andrewdcampbell/doc_scanner/blob/master/sample_images/dollar_bill.JPG" width="350"> <img src="https://github.com/andrewdcampbell/doc_scanner/blob/master/output/dollar_bill.JPG" width="350">
+<img src="sample_images/dollar_bill.JPG" width="350"> <img src="sample_output/dollar_bill.JPG" width="350">
 
 
 ### Usage

diff --git a/output/cell_pic.jpg b/output/cell_pic.jpg
diff --git a/output/chart.JPG b/output/chart.JPG
diff --git a/output/desk.JPG b/output/desk.JPG
diff --git a/output/dollar_bill.JPG b/output/dollar_bill.JPG
diff --git a/output/math_cheat_sheet.JPG b/output/math_cheat_sheet.JPG
diff --git a/output/notepad.JPG b/output/notepad.JPG
diff --git a/output/receipt.jpg b/output/receipt.jpg
diff --git a/output/tax.jpeg b/output/tax.jpeg
diff --git a/before_after.gif → readme_media/before_after.gif b/before_after.gif → readme_media/before_after.gif
diff --git a/ui.gif → readme_media/ui.gif b/ui.gif → readme_media/ui.gif
diff --git a/sample_output/cell_pic.jpg b/sample_output/cell_pic.jpg
diff --git a/sample_output/chart.JPG b/sample_output/chart.JPG
diff --git a/sample_output/desk.JPG b/sample_output/desk.JPG
diff --git a/sample_output/dollar_bill.JPG b/sample_output/dollar_bill.JPG
diff --git a/sample_output/math_cheat_sheet.JPG b/sample_output/math_cheat_sheet.JPG
diff --git a/sample_output/notepad.JPG b/sample_output/notepad.JPG
diff --git a/sample_output/receipt.jpg b/sample_output/receipt.jpg
diff --git a/sample_output/tax.jpeg b/sample_output/tax.jpeg
diff --git a/scan.py b/scan.py
@@ -25,7 +25,7 @@
 class DocScanner(object):
     """An image scanner"""
 
-    def __init__(self, interactive=False, MIN_QUAD_AREA_RATIO=0.25, MAX_QUAD_ANGLE_RANGE=40):
+    def __init__(self, interactive=False, color_preserve=False, MIN_QUAD_AREA_RATIO=0.25, MAX_QUAD_ANGLE_RANGE=40):
         """
         Args:
             interactive (boolean): If True, user can adjust screen contour before
@@ -37,6 +37,7 @@ def __init__(self, interactive=False, MIN_QUAD_AREA_RATIO=0.25, MAX_QUAD_ANGLE_R
                 of its interior angles exceeds MAX_QUAD_ANGLE_RANGE. Defaults to 40.
         """        
         self.interactive = interactive
+        self.color_preserve = color_preserve
         self.MIN_QUAD_AREA_RATIO = MIN_QUAD_AREA_RATIO
         self.MAX_QUAD_ANGLE_RANGE = MAX_QUAD_ANGLE_RANGE        
 
@@ -285,21 +286,24 @@ def scan(self, image_path):
             screenCnt = self.interactive_get_contour(screenCnt, rescaled_image)
 
         # apply the perspective transformation
-        warped = transform.four_point_transform(orig, screenCnt * ratio)
+        output = transform.four_point_transform(orig, screenCnt * ratio) #warped
 
-        # convert the warped image to grayscale
-        gray = cv2.cvtColor(warped, cv2.COLOR_BGR2GRAY)
+        if self.color_preserve == False:
 
-        # sharpen image
-        sharpen = cv2.GaussianBlur(gray, (0,0), 3)
-        sharpen = cv2.addWeighted(gray, 1.5, sharpen, -0.5, 0)
+            # convert the warped image to grayscale
+            gray = cv2.cvtColor(output, cv2.COLOR_BGR2GRAY)
+
+            # sharpen image
+            sharpen = cv2.GaussianBlur(gray, (0,0), 3)
+            sharpen = cv2.addWeighted(gray, 1.5, sharpen, -0.5, 0)
+
+            # apply adaptive threshold to get black and white effect
+            output = cv2.adaptiveThreshold(sharpen, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 21, 15) #thresh
 
-        # apply adaptive threshold to get black and white effect
-        thresh = cv2.adaptiveThreshold(sharpen, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 21, 15)
 
         # save the transformed image
         basename = os.path.basename(image_path)
-        cv2.imwrite(OUTPUT_DIR + '/' + basename, thresh)
+        cv2.imwrite(OUTPUT_DIR + '/' + basename, output)
         print("Proccessed " + basename)
 
 
@@ -310,13 +314,17 @@ def scan(self, image_path):
     group.add_argument("--image", help="Path to single image to be scanned")
     ap.add_argument("-i", action='store_true',
         help = "Flag for manually verifying and/or setting document corners")
+    ap.add_argument("-c", action='store_true',
+        help = "Flag to preserve color on the cropped images")
 
     args = vars(ap.parse_args())
     im_dir = args["images"]
     im_file_path = args["image"]
     interactive_mode = args["i"]
+    color_preserve_mode = args["c"]
 
-    scanner = DocScanner(interactive_mode)
+    scanner = DocScanner(interactive=interactive_mode,
+                            color_preserve=color_preserve_mode)
 
     valid_formats = [".jpg", ".jpeg", ".jp2", ".png", ".bmp", ".tiff", ".tif"]