mmdetection faster-rcnn yolo

mmdetection安裝過程中依靠https://github.com/open-mmlab/mmdetection/blob/master/docs/get_started.md

然后在安裝第三步Install mmcv-full時，發現自己的cuda是10.1的，然后pytorch是1.7.1的然后就用了這條命令

pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu101/torch1.7.1/index.html

實際上是錯誤的，但是沒有報錯，他就直接給你按了一個最新版本的mmcv-full，這和我想要的是不一樣的，我要的是

cuda10.1，pytorch1.7.1的。這是因為這個地址是不存在的

https://download.openmmlab.com/mmcv/dist/cu101/torch1.7.1/index.html

想要得到cuda10.1，pytorch1.7.1的mmcv-full，就需要用

https://download.openmmlab.com/mmcv/dist/cu101/torch1.7.0/index.html
因為這個地址是存在的

這是從mmcv-full的參考表格里知道的，連接是這個https://mmcv.readthedocs.io/en/latest/#install-with-pip

表格是這個，這里的torch1.7就指的是torch1.7.0，在那個http地址上既不能寫成torch1.7，也不能寫成torch1.7.1

然后第三步重要選一種mmcv的安裝方式就行

然后YOLOV3就按下面的弄，然后有一個'`cfg` or `default_args` must contain the key "type"的問題

是因為yolov3_d53_mstrain-608_273e_coco.py文件里的runner = dict(max_epochs=300)這條語句錯了，這是一個還未修正的BUG，

應該改為runner = dict(type='EpochBasedRunner', max_epochs=300)，也就是加上type='EpochBasedRunner'

 

在YOLOtest是，需要用checkpoint的絕對路徑，用相對路徑就會出做錯，不知道為什么，明顯是個BUG

 

 

 

d=====(￣▽￣*)b 我是小小搬運工！站在各位巨人的肩膀上完成噠~~~

哇，再次撒花花~~~

安裝過程

項目地址：https://github.com/open-mmlab/mmdetection

安裝細節

環境配置

python 3.7 pytorch 1.6.0 torchvision 0.7.0 cuda 10.2

• conda create -n mmdetection python=3.7
• git clone https://github.com/open-mmlab/mmdetection.git
• conda install pytorch1.6.0 torchvision0.7.0 cudatoolkit=10.2 -c pytorch

https://github.com/open-mmlab/mmcv

根據這個下載對應的mmcv

• pip install mmcv-full==1.2.2 -f https://download.openmmlab.com/mmcv/dist/cu102/torch1.6.0/index.html

• pip install -r /home/lhh/workspace/AnacondaProjects/mmdetection/mmdetection/requirements/build.txt
  (后面加清華源可能快些，沒有嘗試)

• 運行一段代碼，成功！
  

使用自己數據集訓練

數據集格式

修改路徑與配置（/home/lhh/workspace/AnacondaProjects/mmdetection/mmdetection/configs/base）都是在這個文件夾下設置的

• dataset中的.py文件設置路徑（用到coco數據集就修改對應的py文件）eg:voc0721.py文件
  

• models文件夾修改對應模型的.py文件設置類別數量
• schedules 文件夾修改.py文件設置epoch
• mmdet/datasets/voc.py設置類別名，如果是1類加逗號
• mmdet/core/evaluation/class_names.py設置類別名

以voc數據集，faster_rcnn為例

• 修改schedule_1x.py文件
  修改最后一行的訓練epoch

• 修改配置文件（/home/lhh/workspace/AnacondaProjects/mmdetection/mmdetection/configs/fast_rcnn）中的fast_rcnn_r50_fpn_1x_coco.py設置配置文件的位置，數據類型的位置
  

• 創建文件夾work_dir保存訓練過程及結果

• 運行（具體需要看train.py文件，需要哪些參數,在tools文件夾下）

比如：python tools/train.py configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py --work-dir work_dir

訓練結果：

map結果繪制

• mmdetection$ python tools/analyze_logs.py plot_curve ./work_dir/20201228_234809.log.json --keys mAP --legend mAP --out mAP.jpg
  之后將訓練過程和結果放在統一文件中，上述路徑會所更改
• 參考鏈接：https://www.cnblogs.com/beeblog72/p/12076562.html

• 同理，loss繪制
• python tools/analyze_logs.py plot_curve ./work_dir/20201228_234809.log.json --keys loss --legend loss --out loss.jpg
  
• acc
• python tools/analyze_logs.py plot_curve ./work_dir/faster_rcnn_r50_fpn_1x_coco/20201228_234809.log.json --keys acc --legend acc --out acc.jpg

測試

參考鏈接：https://blog.csdn.net/zxfhahaha/article/details/103754467
注 由於test.py文件只對coco數據集進行eval，所以先用test.py文件生成pkl文件，再用eval_metric.py文件進行計算mAP

• python tools/test.py configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py work_dir/latest.pth --out results.pkl
  –out后面可以加路徑，不然直接生成再項目根路徑下
• python tools/eval_metric.py configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py result.pkl --eval=mAP
  使用pkl文件計算每個類的AP

測試結果

撒花花~~~

以coco數據集，yolov3模型為例

數據集格式：

train2017文件中存放的是圖片
annotations存放的是：

先准備三個相同voc格式的數據集，里面分別存放train、test 和val

• 將某一個txt文本中的數字存的是圖片的名字，要把這些名字的圖片保存到另一個文件夾中

from PIL import Image
f3 = open("F:/dataDB/precoco/val/ImageSets/Main/val.txt",'r') #test文件所在路徑
for line2 in f3.readlines():
    line3=line2[:-1] #讀取每行去掉后四位的數
    im = Image.open('H:/make_data/AB/Images02/{}.jpg'.format(line3))#打開改路徑下的line3記錄的的文件名
    im.save('F:/dataDB/precoco/val/JPEGImages/{}.jpg'.format(line3)) #把文件夾中指定的文件名稱的圖片另存到該路徑下
f3.close()

 

• 將某一個txt文本中的數字存的是圖片的名字，要把這些名字的圖片的xml保存到另一個文件夾中

# -*- coding: UTF-8 -*- 
#!/usr/bin/env python
import sys
import re
import numpy as np

import shutil


data = []
for line in open("F:/dataDB/precoco/val/ImageSets/Main/val.txt", "r"):  # 設置文件對象並讀取每一行文件
    data.append(line)
for a in data:
  #print(a)
  line3=a[:-1] #讀取每行去掉后四位的數,本人使用的格式為000001.jpg，即去掉.jpg
  #print('line3', line3)
  line4 = line3 + '.xml'
  print(line4)
  oldname = r'H:/make_data/AB/Anotations02/{}'.format(line4)
  #print('old', oldname)
  newname = r'F:/dataDB/precoco/val/Annotations/{}'.format(line4)
  #print('new', newname)
  shutil.copyfile(oldname, newname) #將需要的文件從oldname復制到newname

　　

• voc 轉coco數據集

import xml.etree.ElementTree as ET
import os
import json
 
coco = dict()
coco['images'] = []
coco['type'] = 'instances'
coco['annotations'] = []
coco['categories'] = []
 
category_set = dict()
image_set = set()
 
category_item_id = -1
image_id = 20180000000
annotation_id = 0
 
 
def addCatItem(name):
    global category_item_id
    category_item = dict()
    category_item['supercategory'] = 'none'
    category_item_id += 1
    category_item['id'] = category_item_id
    category_item['name'] = name
    coco['categories'].append(category_item)
    category_set[name] = category_item_id
    return category_item_id
 
 
def addImgItem(file_name, size):
    global image_id
    if file_name is None:
        raise Exception('Could not find filename tag in xml file.')
    if size['width'] is None:
        raise Exception('Could not find width tag in xml file.')
    if size['height'] is None:
        raise Exception('Could not find height tag in xml file.')
    image_id += 1
    image_item = dict()
    image_item['id'] = image_id
    image_item['file_name'] = file_name
    image_item['width'] = size['width']
    image_item['height'] = size['height']
    coco['images'].append(image_item)
    image_set.add(file_name)
    return image_id
 
 
def addAnnoItem(object_name, image_id, category_id, bbox):
    global annotation_id
    annotation_item = dict()
    annotation_item['segmentation'] = []
    seg = []
    # bbox[] is x,y,w,h
    # left_top
    seg.append(bbox[0])
    seg.append(bbox[1])
    # left_bottom
    seg.append(bbox[0])
    seg.append(bbox[1] + bbox[3])
    # right_bottom
    seg.append(bbox[0] + bbox[2])
    seg.append(bbox[1] + bbox[3])
    # right_top
    seg.append(bbox[0] + bbox[2])
    seg.append(bbox[1])
 
    annotation_item['segmentation'].append(seg)
 
    annotation_item['area'] = bbox[2] * bbox[3]
    annotation_item['iscrowd'] = 0
    annotation_item['ignore'] = 0
    annotation_item['image_id'] = image_id
    annotation_item['bbox'] = bbox
    annotation_item['category_id'] = category_id
    annotation_id += 1
    annotation_item['id'] = annotation_id
    coco['annotations'].append(annotation_item)
 
 
def parseXmlFiles(xml_path):
    for f in os.listdir(xml_path):
        if not f.endswith('.xml'):
            continue
 
        bndbox = dict()
        size = dict()
        current_image_id = None
        current_category_id = None
        file_name = None
        size['width'] = None
        size['height'] = None
        size['depth'] = None
 
        xml_file = os.path.join(xml_path, f)
        print(xml_file)
 
        tree = ET.parse(xml_file)
        root = tree.getroot()
        if root.tag != 'annotation':
            raise Exception('pascal voc xml root element should be annotation, rather than {}'.format(root.tag))
 
        # elem is <folder>, <filename>, <size>, <object>
        for elem in root:
            current_parent = elem.tag
            current_sub = None
            object_name = None
 
            if elem.tag == 'folder':
                continue
 
            if elem.tag == 'filename':
                file_name = elem.text
                if file_name in category_set:
                    raise Exception('file_name duplicated')
 
            # add img item only after parse <size> tag
            elif current_image_id is None and file_name is not None and size['width'] is not None:
                if file_name not in image_set:
                    current_image_id = addImgItem(file_name, size)
                    print('add image with {} and {}'.format(file_name, size))
                else:
                    raise Exception('duplicated image: {}'.format(file_name))
                    # subelem is <width>, <height>, <depth>, <name>, <bndbox>
            for subelem in elem:
                bndbox['xmin'] = None
                bndbox['xmax'] = None
                bndbox['ymin'] = None
                bndbox['ymax'] = None
 
                current_sub = subelem.tag
                if current_parent == 'object' and subelem.tag == 'name':
                    object_name = subelem.text
                    if object_name not in category_set:
                        current_category_id = addCatItem(object_name)
                    else:
                        current_category_id = category_set[object_name]
 
                elif current_parent == 'size':
                    if size[subelem.tag] is not None:
                        raise Exception('xml structure broken at size tag.')
                    size[subelem.tag] = int(subelem.text)
 
                # option is <xmin>, <ymin>, <xmax>, <ymax>, when subelem is <bndbox>
                for option in subelem:
                    if current_sub == 'bndbox':
                        if bndbox[option.tag] is not None:
                            raise Exception('xml structure corrupted at bndbox tag.')
                        bndbox[option.tag] = int(option.text)
 
                # only after parse the <object> tag
                if bndbox['xmin'] is not None:
                    if object_name is None:
                        raise Exception('xml structure broken at bndbox tag')
                    if current_image_id is None:
                        raise Exception('xml structure broken at bndbox tag')
                    if current_category_id is None:
                        raise Exception('xml structure broken at bndbox tag')
                    bbox = []
                    # x
                    bbox.append(bndbox['xmin'])
                    # y
                    bbox.append(bndbox['ymin'])
                    # w
                    bbox.append(bndbox['xmax'] - bndbox['xmin'])
                    # h
                    bbox.append(bndbox['ymax'] - bndbox['ymin'])
                    print('add annotation with {},{},{},{}'.format(object_name, current_image_id, current_category_id,
                                                                   bbox))
                    addAnnoItem(object_name, current_image_id, current_category_id, bbox)
 
 
if __name__ == '__main__':
    xml_path = 'F:/dataDB/precoco/val/Annotations'    # 這是xml文件所在的地址
    json_file = 'F:/dataDB/precoco/val/ImageSets/val.json'                                     # 這是你要生成的json文件                        
    parseXmlFiles(xml_path)                                       # 只需要改動這兩個參數就行了
    json.dump(coco, open(json_file, 'w'))

　

參考鏈接：
https://blog.csdn.net/weixin_41765699/article/details/100124689

修改過程

• 修改configs/_ base_/datasets文件下的coco_detection.py文件
• coco.py文件的類別名
• class_names.py文件的類別名

錯誤：
之前使用的yolov3…py文件有問題，沒有類別數（num_classes），自己還一直死鑽。。。。
換成如下圖所示：

• 運行代碼：
• python tools/train.py configs/yolo/yolov3_d53_mstrain-608_273e_coco.py --work-dir work_dir/yolov3_d53_320_273e_coco
• 運行成功！

撒花花~~~

測試

• python tools/test.py configs/yolo/yolov3_d53_mstrain-608_273e_coco.py work_dir/yolov3_d53_mstrain-608_273e_coco/latest.pth --out result.pkl --eval bbox