Add files via upload

Author: abdullahbas

Albumentations/albumentationsAugmentation.py

@@ -0,0 +1,118 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Sun Feb 14 15:26:57 2021
+
+@author: trabz
+"""
+import albumentations as A
+
+import torch 
+
+
+from xmlParser import Parser
+import glob
+import numpy as np
+from matplotlib import pyplot as plt
+import cv2
+import torchvision.transforms as transforms
+
+
+class dataloader():
+    
+    def __init__(self,path,transform=None):
+        
+        self.paths=glob.glob(path)
+        self.transform=transform
+        
+        
+    def __len__(self):
+        return len(self.paths)
+    
+    def __getitem__(self,idx):
+        
+        annotation=(Parser.myType(self.paths[idx],idx,classes=['bird','zebra']))
+        image=plt.imread(self.paths[annotation['image_id']].replace('xml','jpg'))
+        
+        if self.transform is not None:
+            
+            augmented = self.transform(image=image, bboxes=annotation['bbox'], labels=annotation['label'])
+
+        
+        return image,augmented,annotation
+
+
+def collate_fn(batch):
+            return tuple(zip(*batch))
+
+
+bbox_params = A.BboxParams(
+  format='pascal_voc', 
+  min_area=1, 
+  min_visibility=0.5, 
+  label_fields=['labels']
+)
+
+aug = A.Compose({
+    #A.Resize(500, 500,p=0.2),
+        A.RGBShift(r_shift_limit=40,g_shift_limit=40,b_shift_limit=40,p=0.04),
+        A.RandomBrightness(p=0.01),
+        A.RandomContrast(p=0.01),
+        A.CLAHE(p=0.02),
+        A.ToGray(p=0.4),
+        A.Blur(blur_limit=8,p=0.1),
+        A.RandomBrightness(p=0.1),
+        A.CenterCrop(100, 100,p=0.01),
+        A.RandomCrop(222, 222,p=0.1),
+        A.HorizontalFlip(p=0.1),
+        A.Rotate(limit=(-90, 90),p=0.2),
+        A.VerticalFlip(p=0.1),
+        A.ShiftScaleRotate(),
+        
+        },bbox_params=bbox_params)
+
+
+path=path= 'Image A*/train/*.xml'
+
+dataset = dataloader(path,aug)
+data_loader = torch.utils.data.DataLoader(
+    dataset, batch_size=1, collate_fn=collate_fn,shuffle=False)
+
+
+
+
+plt.figure(figsize=(12,12))
+for idx,(image,imgO,result) in enumerate(data_loader):
+    
+    
+    imgA=imgO[0]['image']
+    image=image[0]
+    for idc,bbox in enumerate(imgO[0]['bboxes']):
+        xmin,ymin,xmax,ymax=bbox
+        xmin,xmax=np.clip([xmin,xmax],5,imgA.shape[0]-5).astype('int')
+        ymin,ymax=np.clip([ymin,ymax],5,imgA.shape[1]-5).astype('int')
+        
+        xpastmin,ypastmin,xpastmax,ypastmax=np.clip((result[0]['bbox'][idc]),0,max(image.shape)-10)
+        
+        imgA=cv2.rectangle(np.array(imgA),((xmin),(ymin)),((xmax),(ymax))
+                           ,color=[0,245,0],thickness=4)
+       
+        image=cv2.rectangle(image,((xpastmin),(ypastmin)),((xpastmax),(ypastmax))
+                           ,color=[112,9,11],thickness=4)
+    
+    
+    
+    
+    sz,wd,_=np.array(image.shape)-np.array((imgA).shape)
+    img2=imgA
+    imgA=np.pad((imgA),((sz//2,sz//2),(wd//2,wd//2),(0,0)))
+    
+    
+    plt.imsave(f'Albumentations/images/{idx}.png',np.hstack(((imgA).astype('uint8'),image)))
+    plt.axis('off')
+    plt.tight_layout()
+    #plt.imshow(np.dstack((img[:,:,2],img[:,:,1],img[:,:,0])))
+
+
+
+
+