faster-rcnn 目标检测 数据集制作

本文的目标是制作目标检测的数据集

使用的工具是 python + opencv

实现目标

1.批量图片重命名,手动框选图片中的目标，将目标框按照一定格式保存到txt中

图片名格式(批量)

.jpg
.jpg
.jpg

目标包围框储存格式

.jpg car
.jpg car
.jpg car
.jpg car
.jpg car
.jpg car    

实现代码

# -*- coding: cp936 -*-
import os
import cv2
import numpy as np

# 当鼠标按下时变为 True
drawing = False
ix,iy = -1,-1
ox,oy = -1,-1
# 创建回调函数
def draw_circle(event,x,y,flags,param):
    global ix,iy,ox,oy,drawing
    # 当按下左键是返回起始位置坐标
    if event==cv2.EVENT_LBUTTONDOWN:
        drawing=True
        ix,iy = x,y
    # 当鼠标左键按下并移动是绘制图形。 event 可以查看移动, flag 查看是否按下
    elif event==cv2.EVENT_MOUSEMOVE and flags==cv2.EVENT_FLAG_LBUTTON:
         if drawing==True:
                cv2.rectangle(image,(ix,iy),(x,y),(0,255,0),-1)
                ox,oy = x,y
    elif event==cv2.EVENT_LBUTTONUP:
        drawing==False

number = 0
jpg = ".jpg"
Image_Path = "./images"
f_wrect = open('images.txt','a')
for file in os.listdir(Image_Path):
    number = number + 1
    #print(number)
    string_number = '%d'%number
    #print(string_number)
    i = len(string_number)
    #print(i)
    while (5 - i) >  0:
        string_number = '' + string_number
        i = i + 1
    newname = string_number + jpg
    old_NamePath = os.path.join(Image_Path,file)
    new_NamePath = os.path.join(Image_Path,newname)
    os.rename(old_NamePath,new_NamePath)
    image = cv2.imread(new_NamePath)
    cv2.namedWindow('image')
    cv2.setMouseCallback('image',draw_circle)
    while(1):
        cv2.imshow('image',image)
        k=cv2.waitKey(1)&0xFF
        if k==27:
            image_rect = newname + ' car '+ '%d'% ix +' '+ '%d'% iy+ ' ' + '%d'% ox + ' ' + '%d'% oy + '\n'
            f_wrect.write(image_rect)
            break
    cv2.destroyWindow('image')