一、python版本

写了个python opencv的小demo,可以通过键盘按下字母s进行采集图像。

功能说明

“N” 新建文件夹 data/ 用来存储图像 “S” 开始采集图像,将采集到的图像放到 data/ 路径下 “Q” 退出窗口

python opencv源码

'''

“N”  新建文件夹 data/  用来存储图像
"S"   开始采集图像,将采集到的图像放到 data/ 路径下
“Q”   退出窗口
'''

import numpy as np  # 数据处理的库 Numpy
import cv2          # 图像处理的库 OpenCv
import os           # 读写文件
import shutil       # 读写文件
from PIL import Image, ImageDraw, ImageFont


# # OpenCv 调用摄像头 / Use camera
cap = cv2.VideoCapture(0)
cap.set(cv2.CAP_PROP_FRAME_WIDTH,1920)
cap.set(cv2.CAP_PROP_FRAME_HEIGHT,1080)

'''
#功能函数,只是用来往图片中显示汉字
#示例 img = cv2ImgAddText(cv2.imread('img1.jpg'), "大家好,我是片天边的云彩", 10, 65, (0, 0, 139), 20)
参数说明:
img:OpenCV图片格式的图片
text:要写入的汉字
left:字符坐标x值
top:字符坐标y值
textColor:字体颜色
:textSize:字体大小
'''
def cv2ImgAddText(img, text, left, top, textColor=(0, 255, 0), textSize=20):
    if (isinstance(img, np.ndarray)):  # 判断是否OpenCV图片类型
        img = Image.fromarray(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
    # 创建一个可以在给定图像上绘图的对象
    draw = ImageDraw.Draw(img)
    # 字体的格式
    fontStyle = ImageFont.truetype(
        "font/simsun.ttc", textSize, encoding="utf-8")
    # 绘制文本
    draw.text((left, top), text, textColor, font=fontStyle)
    # 转换回OpenCV格式
    return cv2.cvtColor(np.asarray(img), cv2.COLOR_RGB2BGR)

# 存储图像的文件夹 
current_dir = ""
# 保存  图像 的路径 
path_photos_from_camera = "data/"

press_n_flag = 0
cnt_ss=0



while cap.isOpened():
    flag, img_rd = cap.read()
    #print(img_rd.shape)

    kk = cv2.waitKey(2)
    # 待会要写的字体 / Font to write
    font = cv2.FONT_ITALIC

    # 4. 按下 'n' 新建存储人脸的文件夹 / press 'n' to create the folders for saving faces
    if kk == ord('N') or kk == ord('n'):
        current_dir = path_photos_from_camera
        #os.makedirs(current_dir)
        if os.path.isdir(current_dir):
            pass
        else:
            os.mkdir(current_dir)
        print('\n')
        print("新建的保存图像的文件夹 / Create folders: ", current_dir)

        press_n_flag = 1        # 已经按下 'n' / have pressed 'n'


    # 5. 按下 's' 保存摄像头中的图像到本地 / Press 's' to save image into local images
    if kk == ord('S') or kk == ord('s'):
        # 检查有没有先按'n'新建文件夹 / check if you have pressed 'n'
        if press_n_flag:
            cnt_ss += 1
            cv2.imwrite(current_dir + "/img_" + str(cnt_ss) + ".jpg", img_rd)
            print("写入本地 / Save into:", str(current_dir) + "/img_face_" + str(cnt_ss) + ".jpg")
        else:
            print("请在按 'S' 之前先按 'N' 来建文件夹 / Please press 'N' before 'S'")


    # 添加说明 / Add some statements
    #cv2.putText(img_rd, "Face Register", (20, 40), font, 1, (0, 255, 0), 1, cv2.LINE_AA)
    img_rd = cv2ImgAddText(img_rd, "图片采集系统", 160, 25, (0, 255,0), 30)
    #cv2.putText(img_rd, "N: Create face folder", (20, 350), font, 0.8, (0, 255, 0), 1, cv2.LINE_AA)
    img_rd = cv2ImgAddText(img_rd, "N: 创建保存图像文件夹", 20, 350, (0, 255, 0), 20)
    #cv2.putText(img_rd, "S: Save current face", (20, 400), font, 0.8, (0, 255, 0), 1, cv2.LINE_AA)
    img_rd = cv2ImgAddText(img_rd, "S: 保存当前图片", 20, 400, (0, 255, 0), 20)
    #cv2.putText(img_rd, "Q: Quit", (20, 450), font, 0.8, (0, 0, 0), 1, cv2.LINE_AA)
    img_rd = cv2ImgAddText(img_rd, "Q: 退出", 20, 450, (0, 255, 0), 20)

    # 6. 按下 'Q' 键退出 / Press 'q' to exit
    if kk == ord('Q') or kk == ord('q'):
        break
    # 如果需要摄像头窗口大小可调 / Uncomment this line if you want the camera window is resizeable
    cv2.namedWindow("camera", 0)
    cv2.imshow("camera", img_rd)

# 释放摄像头 / Release camera and destroy all windows
cap.release()
cv2.destroyAllWindows()

效果图

安装相关库

windows安装

pip install pillow

tx2/linux/…

sudo apt-get install python3-pillow

二、c语言版本

c语言源码

/*****************************************************
2021.5.18:按键采集图像
******************************************************/
#include "opencv2/core/core.hpp"    
#include "opencv2/imgproc/imgproc.hpp"    
#include "opencv2/calib3d/calib3d.hpp"    
#include "opencv2/highgui/highgui.hpp"    
#include <iostream>    
#include <fstream>    

using namespace cv;
using namespace std;

#define SRC_WIDTH  1920
#define SRC_HEIGHT 1080

int main()
{
	//测试视频
	VideoCapture capture;
	capture.open(1);
	//capture.open("v4l2src device=/dev/video4 ! video/x-raw,width=1920,height=1020,framerate=30/1 ! videoconvert ! appsink");
	if (!capture.isOpened())
	{
		printf("文件打开失败");
	}
	capture.set(CAP_PROP_FRAME_WIDTH, SRC_WIDTH);        //设置宽度
	capture.set(CAP_PROP_FRAME_HEIGHT, SRC_HEIGHT);  //设置长度
	Mat frame;
	int n = 0;
	char* cstr = new char[120];
	while (true)
	{
		
		capture >> frame;
		if (frame.data == NULL)
		{
			printf("Image is empty\n");
			//writer.write(frame);
			break;
			//continue;
		}
		char kk=waitKey(2);
		if (kk == 'S' || kk == 's')
		{

			sprintf(cstr, "%s%d%s", "caliberation/", n++, ".jpg");
			imwrite(cstr, frame);
			printf("保存了图片\n");

		}

		
		namedWindow("111", 0);//参数为零,则可以自由拖动
		imshow("111", frame);
		waitKey(2);
	}

	return 0;

}

效果图

声明:有的资源均来自网络转载,版权归原作者所有,如有侵犯到您的权益 请联系邮箱:312334557@qq.com 我们将配合处理!

原文地址:《python opencv通过按键采集图片源码》发布于2022-03-31 04:21:06