add overexposure image generator

.gitignore

@@ -2,12 +2,11 @@
 .idea/
 XCloud/__pycache__/
 cv/__pycache__/
-cv/static/FaceUpload/
 dm/__pycache__/
 nlp/__pycache__/
 __pycache__/
 pyWeb/
 research/imgcensor/model/
 cv/model/
 cv/static/ImgCensorUpload/
-cv/static/SkinUpload/
+cv/static

README.md

@@ -237,7 +237,7 @@ with open("/path/to/test.jpg", mode='rb') as f:
 
 ## Note
 * XCloud is freely accessible for everyone, you can email me AT 
-**[email protected]** to inquiry tech support.
+**[email protected]** to inquire tech support.
 * **Please ensure that your machine has a strong GPU equipment**.
 * For [XCloud in Java](https://github.com/lucasxlu/CVLH.git), please refer to [CVLH](https://github.com/lucasxlu/CVLH.git) for more details! 
 * Technical details can be read from our [Technical Report](https://arxiv.org/abs/1912.10344). 

research/iqa/README.md

@@ -21,6 +21,10 @@ the deep models to satisfy our requirement. You can also set the last output neu
 
 - I replace the input channel as RGB instead of Gray-scale, since I find RGB input improves accuracy, I also add BatchNorm as a standard component as in SOTA CNN architecture.
 
+- I also provide related code for [blur image generation](blur_creator.py), 
+[warp transformation](./lean_pose_creator.py) and [overexposure image 
+generation](./overexposure_creator.py).
+
 ## Performance Evaluation
 I adopt these models mentioned above for ``exposure/edge recognition`` in ``product recognition project`` to reject unqualified images. The performance is listed as follows, you can train your own model with the code provided within this module.
 
@@ -52,3 +56,6 @@ I adopt these models mentioned above for ``exposure/edge recognition`` in ``prod
 3. Talebi, Hossein, and Peyman Milanfar. ["Nima: Neural image assessment."](https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8352823) IEEE Transactions on Image Processing 27.8 (2018): 3998-4011.
 4. Bosse S, Maniry D, Wiegand T, et al. [A deep neural network for image quality assessment](http://iphome.hhi.de/samek/pdf/BosICIP16.pdf)[C]//2016 IEEE International Conference on Image Processing (ICIP). IEEE, 2016: 3773-3777.
 5. Bianco S, Celona L, Napoletano P, et al. [On the use of deep learning for blind image quality assessment](https://arxiv.org/pdf/1602.05531.pdf)[J]. Signal, Image and Video Processing, 2018, 12(2): 355-362.
+6. Bansal, Raghav, Gaurav Raj, and Tanupriya Choudhury. ["Blur image 
+detection using Laplacian operator and Open-CV."](https://ieeexplore.ieee.org/abstract/document/7894491/) 2016 International 
+Conference System Modeling & Advancement in Research Trends (SMART). IEEE, 2016.

research/iqa/overexposure_creator.py

@@ -0,0 +1,56 @@
+# generate over samples from clear images
+# author: @LucasX
+import argparse
+import os
+import random
+import numpy as np
+
+import cv2
+
+parser = argparse.ArgumentParser()
+parser.add_argument('-orig_dir', type=str,
+                    default='C:/Users/LucasX/Desktop/ShelfExposure/normal')
+parser.add_argument('-outpur_dir', type=str,
+                    default='C:/Users/LucasX/Desktop/ShelfExposure/exposure')
+parser.add_argument('-alpha', type=float, default=2.0)
+parser.add_argument('-beta', type=float, default=0.0)
+parser.add_argument('-show', type=bool, default=False)
+args = vars(parser.parse_args())
+
+print('-' * 100)
+for key, value in args.items():
+    print('%s = %s' % (key, value))
+print('-' * 100)
+
+
+def modify_img_saturation(img_f):
+    """
+    modify image saturation to imitate overexposure effect
+    :param img_f:
+    :return:
+    """
+    if img_f.endswith('.jpg') or img_f.endswith('.png') or img_f.endswith('.jpeg'):
+        if not os.path.exists(args['outpur_dir']):
+            os.makedirs(args['outpur_dir'])
+        image = cv2.imread(img_f)
+
+        overexposure_image = np.zeros(image.shape, image.dtype)
+
+        for y in range(image.shape[0]):
+            for x in range(image.shape[1]):
+                for c in range(image.shape[2]):
+                    # alpha = args['alpha']
+                    alpha = random.randint(2, 10)
+                    overexposure_image[y, x, c] = np.clip(alpha * image[y, x, c] + args['beta'], 0, 255)
+
+        if args['show'] and overexposure_image is not None:
+            cv2.imshow('overexposure_image', overexposure_image)
+            cv2.waitKey(0)
+            cv2.destroyAllWindows()
+
+        cv2.imwrite(os.path.join(args['outpur_dir'], os.path.basename(img_f)), overexposure_image)
+
+
+if __name__ == '__main__':
+    for img_f in os.listdir(args['orig_dir']):
+        modify_img_saturation(os.path.join(args['orig_dir'], img_f))