⑤ Darkflow : YOLO의 Tensorflow 버전

YOLO의 Tensorflow버전인 Darkflow를 사용해보겠습니다.

Darkflow란?

---

기본 YOLO는 C를 기반으로 작성되었습니다. 이를 Tensorflow으로 작성된것이 darkflow입니다.

YOLO에 대한 설명과 사용법은 ☞여기☜에 작성했습니다.

darkflow 깃허브 : https://github.com/thtrieu/darkflow

실행 환경

---

UBUNTU	16.04
CUDA_VERSION	9.1.85
CUDNN_VERSION	7.1.2.21
OpenCV	3
Python	3
Tensorflow	1.0

labeling 

---

from lxml import etree
from PIL import Image
import csv
import os
 
IMG_PATH = "D:/__Project__/darkflow-master/data/train_traffic_light"
fw = os.listdir(IMG_PATH)
# path of save xml file
save_path = 'D:/__Project__/darkflow-master/data/annotations/'
 
# txt_folder is txt file root that using darknet rectbox
txt_folder = 'D:/__Project__/darkflow-master/data/train_traffic_light_TXT'
 
# edit ypur label set
labels = ['traffic light']
 
def csvread(fn):
    with open(fn, 'r') as csvfile:
        list_arr = []
        reader = csv.reader(csvfile, delimiter=' ')
 
        for row in reader:
            list_arr.append(row)
    return list_arr
 
 
def convert_label(txt_file):
    if((txt_file[0]) == str(0)):
        label = 'traffic light'
 
    return label
 
# core code = convert the yolo txt file to the x_min,x_max...
 
 
def extract_coor(txt_file, img_width, img_height):
    x_rect_mid = float(txt_file[1])
    y_rect_mid = float(txt_file[2])
    width_rect = float(txt_file[3])
    height_rect = float(txt_file[4])
 
    x_min_rect = ((2 * x_rect_mid * img_width) - (width_rect * img_width)) / 2
    x_max_rect = ((2 * x_rect_mid * img_width) + (width_rect * img_width)) / 2
    y_min_rect = ((2 * y_rect_mid * img_height) -
                  (height_rect * img_height)) / 2
    y_max_rect = ((2 * y_rect_mid * img_height) +
                  (height_rect * img_height)) / 2
 
    return x_min_rect, x_max_rect, y_min_rect, y_max_rect
 
 
for line in fw:
    root = etree.Element("annotation")
 
    # try debug to check your path
    img_style = IMG_PATH.split('/')[-1]
    img_name = line
    image_info = IMG_PATH + "/" + line
    img_txt_root = txt_folder + "/" + line[:-4]
    txt = ".txt"
 
    txt_path = img_txt_root + txt
    txt_file = csvread(txt_path)
    ######################################
 
    # read the image  information
    img_size = Image.open(image_info).size
 
    img_width = img_size[0]
    img_height = img_size[1]
    img_depth = Image.open(image_info).layers
    ######################################
 
    folder = etree.Element("folder")
    folder.text = "%s" % (img_style)
 
    filename = etree.Element("filename")
    filename.text = "%s" % (img_name)
 
    path = etree.Element("path")
    path.text = "%s" % (IMG_PATH)
 
    source = etree.Element("source")
    ##################source - element##################
    source_database = etree.SubElement(source, "database")
    source_database.text = "Unknown"
    ####################################################
 
    size = etree.Element("size")
    ####################size - element##################
    image_width = etree.SubElement(size, "width")
    image_width.text = "%d" % (img_width)
 
    image_height = etree.SubElement(size, "height")
    image_height.text = "%d" % (img_height)
 
    image_depth = etree.SubElement(size, "depth")
    image_depth.text = "%d" % (img_depth)
    ####################################################
 
    segmented = etree.Element("segmented")
    segmented.text = "0"
 
    root.append(folder)
    root.append(filename)
    root.append(path)
    root.append(source)
    root.append(size)
    root.append(segmented)
 
    for ii in range(len(txt_file)):
 
        label = convert_label(txt_file[ii][0])
        x_min_rect, x_max_rect, y_min_rect, y_max_rect = extract_coor(
            txt_file[ii], img_width, img_height)
 
        object = etree.Element("object")
        ####################object - element##################
        name = etree.SubElement(object, "name")
        name.text = "%s" % (label)
 
        pose = etree.SubElement(object, "pose")
        pose.text = "Unspecified"
 
        truncated = etree.SubElement(object, "truncated")
        truncated.text = "0"
 
        difficult = etree.SubElement(object, "difficult")
        difficult.text = "0"
 
        bndbox = etree.SubElement(object, "bndbox")
        #####sub_sub########
        xmin = etree.SubElement(bndbox, "xmin")
        xmin.text = "%d" % (x_min_rect)
        ymin = etree.SubElement(bndbox, "ymin")
        ymin.text = "%d" % (y_min_rect)
        xmax = etree.SubElement(bndbox, "xmax")
        xmax.text = "%d" % (x_max_rect)
        ymax = etree.SubElement(bndbox, "ymax")
        ymax.text = "%d" % (y_max_rect)
        #####sub_sub########
 
        root.append(object)
        ####################################################
 
    file_output = etree.tostring(root, pretty_print=True, encoding='UTF-8')
    # print(file_output.decode('utf-8'))
    ff = open('%s%s.xml' % (save_path, img_name[:-4]), 'w', encoding="utf-8")
    ff.write(file_output.decode('utf-8'))

사용법

---

pip가 등록된 패키지가 아니므로 git을 사용해 다운받아야합니다.

https://github.com/thtrieu/darkflow.git
 
cd darkflow
 
python3 setup.py build_ext --inplace
 
pip install .

현재 yolov3는 지원하지 않습니다.  Issue #665

1. label.txt 수정

./label.txt의 내용을 해당 classes로 수정한다.

diseased

2. ./cfg/yolo.cfg 수정

사용할 cfg파일을 복사해 다른이름으로 저장

convolutional의 filter = (archors+class)*num 로 변경

yolo 레이어의 classes를 수정

[convolutional]
size=1
stride=1
pad=1
filters=425
activation=linear
 
 
[region]
anchors =  0.57273, 0.677385, 1.87446, 2.06253, 3.33843, 5.47434, 7.88282, 3.52778, 9.77052, 9.16828
bias_match=1
classes=80
coords=4
num=5
softmax=1
jitter=.3
rescore=1

Training

---

(1) CUDA_VISIBLE_DEVICES=2 python 자체로 GPU사용해서 돌리기

(2) 학습할 때 --train 꼭 사용

(3) --dataset ./data/corn_img/ 에 이미지 사진 폴더 위치(.jpg .png)

(4) --annotation ./data/annotation_corn/에 annotation 데이터 넣기 (.xml)

(5) 나머지는 parameter 참고

CUDA_VISIBLE_DEVICES=2 python3 flow \
--model ./cfg/my-yolo.cfg \
--labels ./labels.txt \
--dataset ./data/corn_img/ \
--annotation ./data/annotation_corn/ \
--train \
--trainer adam \
--batch 5 \
--epoch 100 \
--save 50 \
--keep 5 \
--lr 1e-04 \
--gpu 0.9 \
--load -1
--summary ./logs \
--gpuName /gpu:2 \
--savepb True \

detect

---

(1) --imgdir ./sample_img/ttt/테스트 이미지 위치 넣기

(2) --threshold 0.5 임계값 설정해서 0.5이상만 표시

python3 flow \
--imgdir ./sample_img/ttt/ \
--model ./cfg/my-tiny-yolo.cfg \
--load -1 \
--batch 1 \
--threshold 0.5 

참고

---

flow --help

Example usage: flow --imgdir sample_img/ --model cfg/yolo.cfg --load bin/yolo.weights
 
Arguments:
  --help, --h, -h  show this super helpful message and exit
  --imgdir         path to testing directory with images
  --binary         path to .weights directory
  --config         path to .cfg directory
  --dataset        path to dataset directory
  --labels         path to labels file
  --backup         path to backup folder
  --summary        path to TensorBoard summaries directory
  --annotation     path to annotation directory
  --threshold      detection threshold
  --model          configuration of choice
  --trainer        training algorithm
  --momentum       applicable for rmsprop and momentum optimizers
  --verbalise      say out loud while building graph
  --train          train the whole net
  --load           how to initialize the net? Either from .weights or a checkpoint, or even from scratch
  --savepb         save net and weight to a .pb file
  --gpu            how much gpu (from 0.0 to 1.0)
  --gpuName        GPU device name
  --lr             learning rate
  --keep           Number of most recent training results to save
  --batch          batch size
  --epoch          number of epoch
  --save           save checkpoint every ? training examples
  --demo           demo on webcam
  --queue          process demo in batch
  --json           Outputs bounding box information in json format.
  --saveVideo      Records video from input video or camera
  --pbLoad         path to .pb protobuf file (metaLoad must also be specified)
  --metaLoad       path to .meta file generated during --savepb that corresponds to .pb file