From 0ecbeeff42b272b2a8973a8c3dde95703f227915 Mon Sep 17 00:00:00 2001
From: salbatron <29922029+salbatron@users.noreply.github.com>
Date: Thu, 17 May 2018 15:22:49 +0200
Subject: [PATCH] Create get_openimages_dataset.py

Script that downloads images with selected classes from OpenImages dataset and create labels in "yolo format"
---
 scripts/get_openimages_dataset.py | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)
 create mode 100644 scripts/get_openimages_dataset.py

diff --git a/scripts/get_openimages_dataset.py b/scripts/get_openimages_dataset.py
new file mode 100644
index 00000000..3508b3ec
--- /dev/null
+++ b/scripts/get_openimages_dataset.py
@@ -0,0 +1,19 @@
+import csv
+import os
+
+#select classes you want to download at https://github.com/openimages/dataset/blob/master/dict.csv
+CLASS_LIST = ('/m/01g317','/m/04yx4')
+img_name = "111111111111"
+
+#download csv from https://storage.googleapis.com/openimages/web/download.html
+with open('path\\train-annotations-bbox.csv', newline='') as csvfile:
+    bboxs = csv.reader(csvfile, delimiter=',', quotechar='|')
+    for bbox in bboxs:
+        if bbox[2] in CLASS_LIST:
+            if img_name != bbox[0]:
+                if not os.path.isfile("destination_path\\%s.jpg"%bbox[0]):
+                    os.system("gsutil cp gs://open-images-dataset/train/%s.jpg destination_path"%bbox[0])
+                    out_file = open("destination_path\\%s.txt"%bbox[0], 'w')
+                    img_name = bbox[0]
+            if img_name == bbox[0]:
+                out_file.write(str(CLASS_LIST.index(bbox[2])) + " " + str(float(bbox[4])+(float(bbox[5])-float(bbox[4]))/2) + " " + str(float(bbox[6])+(float(bbox[7])-float(bbox[6]))/2)+ " " + str(float(bbox[5])-float(bbox[4])) + " " + str(float(bbox[7])-float(bbox[6])) + '\n')