对于部分初学者,偶尔会想在代码中查看图片的位深度,一般我们会用Mat.type()来获得类型,但是默认输出的是整型数字,不便于理解,可能还需要自己对照宏定义去查看,这里有一段代码可以实现将整型转为宏定义位深度
string Type2String(int type)
{
string strType;
uchar depth = type & CV_MAT_DEPTH_MASK;
uchar chans = 1 + (type >> CV_CN_SHIFT);
switch (depth)
{
case CV_8U:
strType = "CV_8U"; break;
case CV_8S:
strType = "CV_8S"; break;
case CV_16U:
strType = "CV_16U"; break;
case CV_16S:
strType = "CV_16S"; break;
case CV_32S:
strType = "CV_32S"; break;
case CV_32F:
strType = "CV_32F"; break;
case CV_64F:
strType = "CV_64F"; break;
default:
strType = "UNKNOWN_TYPE"; break;
}
strType += "C";
strType += (chans + '0');
return strType;
}
比如,使用imread读取图片,如果flags参数不填,默认以彩色模式读取并转为8位,那么类型就是CV_8UC3, 如果flags设置位0,则以灰度模式读取,类型位CV_8UC1。但是,有时候我们会读取一些16位或32位的图片,这时候就需要设置flags参数为 IMREAD_UNCHANGED,这样读取的图片位深度才会被正确显示,如下:
Mat img = imread("1.tif", IMREAD_UNCHANGED);
cout << img.type() << endl;
cout << Type2String(img.type()) << endl;
flags参数其他含义可以参考定义说明查看:
完整代码:
#include<iostream>
#include<opencv2/opencv.hpp>
using namespace std;
using namespace cv;
string Type2String(int type)
{
string strType;
uchar depth = type & CV_MAT_DEPTH_MASK;
uchar chans = 1 + (type >> CV_CN_SHIFT);
switch (depth)
{
case CV_8U:
strType = "CV_8U"; break;
case CV_8S:
strType = "CV_8S"; break;
case CV_16U:
strType = "CV_16U"; break;
case CV_16S:
strType = "CV_16S"; break;
case CV_32S:
strType = "CV_32S"; break;
case CV_32F:
strType = "CV_32F"; break;
case CV_64F:
strType = "CV_64F"; break;
default:
strType = "UNKNOWN_TYPE"; break;
}
strType += "C";
strType += (chans + '0');
return strType;
}
void main()
{
Mat img = imread("1.tif", IMREAD_UNCHANGED);
cout << img.type() << endl;
cout << Type2String(img.type()) << endl;
}
本文分享自 OpenCV与AI深度学习 微信公众号,前往查看
如有侵权,请联系 cloudcommunity@tencent.com 删除。
本文参与 腾讯云自媒体同步曝光计划 ,欢迎热爱写作的你一起参与!