* 目标检测数据集支持单日期格式新增,支持所有日期格式新增

Author: jadehh

CONTRIBUTING.md

@@ -1,12 +1,17 @@
 ### 更新日志
 
-#### DatasetToolsV1.1.5 - 2023-10-13
-* 支持新增数据集
+#### DatasetToolsV1.1.6 - 2024-02-26
+* 目标检测数据集支持单日期格式新增,支持所有日期格式新增
 ---
 
 <details onclose>
 <summary>查看更多更新日志</summary>
 
+
+#### DatasetToolsV1.1.5 - 2023-10-13
+* 支持新增数据集
+---
+
 #### DatasetToolsV1.1.4 - 2023-09-27
 * 降低箱型校验标准
 ---

dataset_tools/jade_create_object_dection_datasets.py

@@ -8,11 +8,18 @@
 # @Desc     : 制作目标检测数据集
 from dataset_tools import *
 import os
-from jade import ProgressBar,GetLastDir
+from jade import ProgressBar,GetLastDir,CreateSavePath
 import shutil
 import random
 import xml.etree.ElementTree as ET
 
+def CreateSavePath(save_path):
+    if os.path.exists(save_path):
+        return save_path
+    else:
+        os.makedirs(save_path)
+        return save_path
+
 def ProcessXml(xml_path):
     # Read the XML annotation file.
     tree = ET.parse(xml_path)
@@ -52,47 +59,64 @@ def ProcessXml(xml_path):
     imagename = GetLastDir(xml_path)[:-4]+'.jpg'
     return imagename,shape, bboxes, labels_text,labels, difficult, truncated
 
-def CreateYearsDatasets(dir,rate=0.95):
+def CreateYearsDatasets(dir,year=None,save_path=None,rate=0.95):
     years = os.listdir(dir)
-    if os.path.exists(os.path.join(dir,"train.txt")):
-        os.remove(os.path.join(dir,"train.txt"))
-    if os.path.exists(os.path.join(dir,"test.txt")):
-        os.remove(os.path.join(dir,"test.txt"))
-    progressBar1 = ProgressBar(len(years))
-    with open(os.path.join(dir,"train.txt"),"w") as f1:
+    if os.path.exists(os.path.join(save_path,"train.txt")):
+        os.remove(os.path.join(save_path,"train.txt"))
+    if os.path.exists(os.path.join(save_path,"test.txt")):
+        os.remove(os.path.join(save_path,"test.txt"))
+    if year is None:
+        progressBar1 = ProgressBar(len(years))
+    else:
+        progressBar1 = ProgressBar(1)
+    if os.path.exists(save_path):
+        pass
+    else:
+        os.makedirs(save_path)
+    if year is None:
         for year in years:
             if os.path.isdir(os.path.join(dir, year)):
-                if os.path.exists(os.path.join(dir,year,DIRECTORY_IMAGES)) and os.path.exists(os.path.join(dir,year,DIRECTORY_ANNOTATIONS)):
-                    CreateVOCDataset(os.path.join(dir,year),year,rate)
-                with open(os.path.join(dir,year,"ImageSets","Main","train.txt")) as f2:
-                    for content in f2.read().split("\n")[:-1]:
-                        f1.write(content+"\n")
+                if os.path.exists(os.path.join(dir, year, DIRECTORY_IMAGES)) and os.path.exists(
+                        os.path.join(dir, year, DIRECTORY_ANNOTATIONS)):
+                    CreateVOCDataset(os.path.join(dir, year), year, save_path, rate)
             progressBar1.update()
-
-    progressbar2 = ProgressBar(len(years))
-    with open(os.path.join(dir, "test.txt"), "w") as f1:
+    else:
+        if os.path.isdir(os.path.join(dir, year)):
+            if os.path.exists(os.path.join(dir, year, DIRECTORY_IMAGES)) and os.path.exists(
+                    os.path.join(dir, year, DIRECTORY_ANNOTATIONS)):
+                CreateVOCDataset(os.path.join(dir, year), year, save_path, rate)
+
+    years = os.listdir(save_path)
+    with open(os.path.join(save_path, "train.txt"), "w") as f1:
+        progressbar2 = ProgressBar(len(years))
         for year in years:
             if os.path.isdir(os.path.join(dir, year)):
-                if os.path.exists(os.path.join(dir,year,DIRECTORY_IMAGES)) and os.path.exists(os.path.join(dir,year,DIRECTORY_ANNOTATIONS)):
-                    CreateVOCDataset(os.path.join(dir,year),year,rate)
-                with open(os.path.join(dir, year, "ImageSets", "Main", "test.txt")) as f2:
+                with open(os.path.join(save_path, year, "ImageSets", "Main", "train.txt")) as f2:
                     for content in f2.read().split("\n")[:-1]:
+                        f1.write(content + "\n")
+        progressbar2.update()
 
+    with open(os.path.join(save_path, "test.txt"), "w") as f1:
+        progressbar2 = ProgressBar(len(years))
+        for year in years:
+            if os.path.isdir(os.path.join(dir, year)):
+                with open(os.path.join(save_path, year, "ImageSets", "Main", "test.txt")) as f2:
+                    for content in f2.read().split("\n")[:-1]:
                         f1.write(content + "\n")
-            progressbar2.update()
-    CreateLabelList(dir)
+        progressbar2.update()
+    CreateLabelList(save_path)
 
 
 ##制作VOC数据集
-def CreateVOCDataset(dir, datasetname,rate=0.95):
+def CreateVOCDataset(dir, datasetname,save_path=None,rate=0.95):
     """
 
     :param dir:
     :param datasetname:
     :param rate:
     :return:
     """
-    root_path = dir
+    root_path = os.path.join(save_path,datasetname)
     dataset_name = datasetname
     Annotations = DIRECTORY_ANNOTATIONS
     JPEGImages = DIRECTORY_IMAGES
@@ -103,8 +127,7 @@ def CreateVOCDataset(dir, datasetname,rate=0.95):
         shutil.rmtree(os.path.join(root_path, "ImageSets", "Main"))
         os.makedirs(os.path.join(root_path, "ImageSets", "Main"))
     Main_path = os.path.join(root_path, "ImageSets", "Main")
-    image_files = os.listdir(os.path.join(root_path, JPEGImages))
-
+    image_files = os.listdir(os.path.join(dir, JPEGImages))
     train_image_files = random.sample(image_files, int(len(image_files) *rate))
     test_image_files = [file for file in image_files if file not in train_image_files]
 
@@ -114,10 +137,14 @@ def CreateVOCDataset(dir, datasetname,rate=0.95):
             image_file = dataset_name + "/" + JPEGImages + "/" + train_image_file
             xml_file = dataset_name + "/" + Annotations + "/" + train_image_file[:-4] + ".xml"
             filename = train_image_file[:-4]
+            save_image_path = CreateSavePath(os.path.join(root_path,DIRECTORY_IMAGES))
+            save_xml_path = CreateSavePath(os.path.join(root_path,DIRECTORY_ANNOTATIONS))
             with open(os.path.join(dir,JPEGImages,train_image_file),"rb") as f2:
                 if len(f2.read()) == 0:
                     pass
                 else:
+                    shutil.copy(os.path.join(dir, JPEGImages, train_image_file), save_image_path)
+                    shutil.copy(os.path.join(dir, DIRECTORY_ANNOTATIONS, train_image_file[:-4] + ".xml"), save_xml_path)
                     f.write(filename + "\n")
             # f.write(image_file + " " + xml_file + "\n")
 
@@ -127,10 +154,14 @@ def CreateVOCDataset(dir, datasetname,rate=0.95):
             image_file = dataset_name + "/" + JPEGImages + "/" + test_image_file
             xml_file = dataset_name + "/" + Annotations + "/" + test_image_file[:-4] + ".xml"
             filename = test_image_file[:-4]
+            save_image_path = CreateSavePath(os.path.join(root_path, DIRECTORY_IMAGES))
+            save_xml_path = CreateSavePath(os.path.join(root_path, DIRECTORY_ANNOTATIONS))
             with open(os.path.join(dir,JPEGImages,test_image_file),"rb") as f2:
                 if len(f2.read()) == 0:
                     pass
                 else:
+                    shutil.copy(os.path.join(dir, JPEGImages, train_image_file), save_image_path)
+                    shutil.copy(os.path.join(dir, DIRECTORY_ANNOTATIONS, train_image_file[:-4] + ".xml"), save_xml_path)
                     f.write(filename + "\n")
             # f.write(image_file + " " + xml_file + "\n")
 
@@ -179,10 +210,10 @@ def CreateLabelList(dir):
                         label_list.append(label)
                 progressbar.update()
         progressBar2 = ProgressBar(len(label_list))
-        for label_name in label_list:
-            with open(os.path.join(dir,"label_list.txt"),"a") as f:
-                f.write(label_name+"\n")
-            progressBar2.update()
+        with open(os.path.join(dir, "label_list.txt"), "wb") as f:
+            for label_name in label_list:
+                f.write((label_name+"\n").encode("utf-8"))
+        progressBar2.update()
     else:
         print("请先制作数据集")
 

test.py

@@ -0,0 +1,12 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+# @File     : test.py
+# @Author   : jade
+# @Date     : 2024/2/27 9:16
+# @Email    : jadehh@1ive.com
+# @Software : Samples
+# @Desc     :
+from dataset_tools.jade_create_object_dection_datasets import *
+
+if __name__ == '__main__':
+    CreateYearsDatasets("F:\数据集\VOC数据集\箱门检测数据集\ContainVOC",year=None,save_path="E:\数据集\VOC数据集\箱门检测数据集\ContainVOC",rate=0.95)