Is there a direct way to compute the column-wise standard deviation for a matrix in opencv? Similar to std in Matlab. I've found one for the mean:
cv::Mat col_mean;
reduce(A, col_mean, 1, CV_REDUCE_AVG);
but I cannot find such a function for the standard deviation.
Here's a quick answer to what you're looking for. I added both the standard deviation and mean for each column. The code can easily be modified for rows.
cv::Mat A = ...; // FILL IN THE DATA FOR YOUR INPUT MATRIX
cv::Mat meanValue, stdValue;
cv::Mat colSTD(1, A.cols, CV_64FC1);
cv::Mat colMEAN(1, A.cols, CV_64FC1);
for (int i = 0; i < A.cols; i++){
cv::meanStdDev(A.col(i), meanValue, stdValue);
colSTD.at<double>(i) = stdValue.at<double>(0);
colMEAN.at<double>(i) = meanValue.at<double>(0);
}
The following is not in a single line,but it is another version without loops:
reduce(A, meanOfEachCol, 0, CV_REDUCE_AVG); // produces single row of columnar means
Mat repColMean;
cv::repeat(meanOfEachCol, rows, 1, repColMean); // repeat mean vector 'rows' times
Mat diffMean = A - repColMean; // get difference
Mat diffMean2 = diffMean.mul(diffMean); // per element square
Mat varMeanF;
cv::reduce(diffMean2, varMeanF, 0, CV_REDUCE_AVG); // sum each column's elements to get single row
Mat stdMeanF;
cv::sqrt(varMeanF, stdMeanF); // get standard deviation
Related
I am using OpenCV 3.1 with VS2012 C++/CLI.
I have stored the result of a finContours call into:
std::vector<std::vector<Point>> Contours;
Thus, Contours[0] is a vector of the contour points of the first contour.
Contours[1] is a vector of the contour points of the second vector, etc.
Now, I want to load one of the contours into a Mat Based on Convert Mat to vector <float> and Vector<float> to mat in opencv I thought something like this would work.
Mat testMat=Mat(Images->Contours[0].size(),2,CV_32FC1);
memcpy(testMat.data,Images->Contours[0].data(),Images->Contours[0].size()*CV_32FC1);
I specified two columns because I each underlying pint must be composed of both an X point and a Y point and each of those should be a float. However, when I access the Mat elements, I can see that the first element is not the underlying data but the total number of contour points.
Any help on the right way to accomplish this appreaciated.
You can do that with:
Mat testMat = Mat(Images->Contours[0]).reshape(1);
Now testMat is of type CV_32SC1, aka of int. If you need float you can:
testMat.convertTo(testMat, CV_32F);
Some more details and variants...
You can simply use the Mat constructor that accepts a std::vector:
vector<Point> v = { {0,1}, {2,3}, {4,5} };
Mat m(v);
With this, you get a 2 channel matrix with the underlying data in v. This means that if you change the value in v, also the values in m change.
v[0].x = 7; // also 'm' changes
If you want a deep copy of the values, so that changes in v are not reflected in m, you can use clone:
Mat m2 = Mat(v).clone();
Your matrices are of type CV_32SC2, i.e. 2 channels matrices of int (because Point uses int. Use Point2f for float). If you want a 2 columns single channel matrix you can use reshape:
Mat m3 = m2.reshape(1);
If you want to convert to float type, you need to use convertTo:
Mat m4;
m2.convertTo(m4, CV_32F);
Here some working code as a proof of concept:
#include <opencv2\opencv.hpp>
#include <vector>
using namespace std;
using namespace cv;
int main()
{
vector<Point> v = { {0,1}, {2,3}, {4,5} };
// changes in v affects m
Mat m(v);
// changes in v doesn't affect m2
Mat m2 = Mat(v).clone();
// m is changed
v[0].x = 7;
// m3 is a 2 columns single channel matrix
Mat m3 = m2.reshape(1);
// m4 is a matrix of floats
Mat m4;
m2.convertTo(m4, CV_32F);
return 0;
}
I try to use cvCalibrateCamera2, but I get error that rotation matrix is not properly defined:
...calibration.cpp:1495: error: (-5) the output array of rotation vectors must be 3-channel 1xn or nx1 array or 1-channel nx3 or nx9 array, where n is the number of views
I have already tried all possibilities from that info but I still get this error.
My code:
CvMat *object_points = cvCreateMat((int)pp.object_points.size(), 1, CV_32FC3);
CvMat *image_points = cvCreateMat((int)pp.image_points.size(), 1, CV_32FC2);
const CvMat point_counts = cvMat((int)pp.point_counts.size(), 1, CV_32SC1, &pp.point_counts[0]);
for (size_t i=0; i<pp.object_points.size(); i++)
{
object_points->data.fl[i*3+0] = (float)pp.object_points[i].x;
object_points->data.fl[i*3+1] = (float)pp.object_points[i].y;
object_points->data.fl[i*3+2] = (float)pp.object_points[i].z;
image_points->data.fl[i*2+0] = (float)pp.image_points[i].x;
image_points->data.fl[i*2+1] = (float)pp.image_points[i].y;
}
CvMat* tempR = cvCreateMat(1, 3, CV_32F);
cvCalibrateCamera2(object_points, image_points, &point_counts,
cvSize(pp.width, pp.height), camera->m_calib_K,
camera->m_calib_D, tempR, &tempData->m_calib_T,
CV_CALIB_USE_INTRINSIC_GUESS)
// camera->calib_T is defined as:
// double m_calib_T_data[3];
// cvMat(3, 1, CV_64F, camera->m_calib_T_data);
I thought that rotation matrix used in cvCalibrateCamera2 should be 1x3 (then I want to use Rodrigues function to get 3x3 matrix) but it doesn't work as any other combination mentioned in error.
Any ideas?
And I use opencv 2.4.0 (maybe there is bug in that method, but for some reasons I can't use later version of opencv)
I think the statement is clear. I am not confident with C# but I know it requires a strong initialization.
The problem in line
CvMat* tempR = cvCreateMat(1, 3, CV_32F);
is that tempR should have a line 1x3 for every N objects point you use. In this sense, the statement becomes clear
...calibration.cpp:1495: error: (-5) the output array of rotation
vectors must be 3-channel 1xn or nx1 array or 1-channel nx3 or nx9
array, where n is the number of views
You must create a tempR like that (more or less, I do not know how to calculate N in C#)
CvMat* tempR = cvCreateMat(N, 3, CV_32F);
Try to extract N from dimensions of object.point.size. If it does not work, try image.point.size
I am trying to figure a way of sorting a 3x3 row into a 9x1.
So i have following:
I want to end up with this:
This is what i end up doing so far:
Rect roi(y-1,x-1,kernel,kernel);
Mat image_roi = image(roi);
Mat image_sort(kernel, kernel, CV_8U);
cv::sort(image_roi, image_sort, CV_SORT_ASCENDING+CV_SORT_EVERY_ROW);
The code is not functional, currently i cannot find any data in my image_sort after its "sorted".
Sure you have single-channel grey level images? Try:
cv::Mat image_sort = cv::Mat::zeros(rect.height, rect.width, rect.type()); // allocated memory
image(roi).copyTo(image_sort); // copy data in image_sorted
std::sort(image_sort.data, image_sort.dataend); // call std::sort
cv::Mat vectorized = image_sort.reshape(1, 1); // reshaped your WxH matrix into a 1x(W*H) vector.
I'm trying to create a PCA model in OpenCV to hold pixel coordinates. As an experiment I have two sets of pixel coordinates that maps out two approximate circles. Each set of coordiantes has 48 x,y pairs. I was experimenting with the following code which reads the coordinates from a file and stores them in a Mat structure. However, I don't think it is right and PCA in openCV seems very poorly covered on the Internet.
Mat m(2, 48, CV_32FC2); // matrix with 2 rows of 48 cols of floats held in two channels
pFile = fopen("data.txt", "r");
for (int i=0; i<48; i++){
int x, y;
fscanf(pFile, "%d%c%c%d%c", &x, &c, &c, &y, &c);
m.at<Vec2f>( 0 , i )[0] = (float)x; // store x in row 0, col i in channel 0
m.at<Vec2f>( 0 , i )[1] = (float)y; // store y in row 0, col i in channel 1
}
for (int i=0; i<48; i++){
int x, y;
fscanf(pFile, "%d%c%c%d%c", &x, &c, &c, &y, &c);
m.at<Vec2f>( 1 , i )[0] = (float)x; // store x in row 1, col i in channel 0
m.at<Vec2f>( 1 , i )[1] = (float)y; // store y in row 1, col i in channel 1
}
PCA pca(m, Mat(), CV_PCA_DATA_AS_ROW, 2); // 2 principle components??? Not sure what to put here e.g. is it 2 for two data sets or 48 for number of elements?
for (int i=0; i<48; i++){
float x = pca.mean.at<Vec2f>(i,0)[0]; //get average x
float y = pca.mean.at<Vec2f>(i,0)[1]; //get average y
printf("\n x=%f, y=%f", x, y);
}
However, this crashes when creating the pca object. I know this is a very basic question but I am a bit lost and was hoping that someone could get me started with pca in open cv.
Perhaps it would be helpful if you described in further detail what you need to use PCA for and what you hope to achieve (output?).
I am fairly sure that the reason your program crashes is because the input Mat is CV_32FC2, when it should be CV_32FC1. You need to reshape your data into 1 dimensional row vectors before using PCA, not knowing what you need I can't say how to reshape your data. (The common application with images is eigenFace which requires an image to be reshaped into a row vector). Additionally you will need to normalize your input data between 0 and 1.
As a further aside, usually you would choose to keep 1 less principal component than the number of input samples because the last principal component is simply orthogonal to the others.
I have worked with opencv PCA before and would like to help further. I would also refer you to this blog: http://www.bytefish.de/blog/pca_in_opencv which helped me get started with PCA in openCV.
In Matlab, If A is a matrix, sum(A) treats the columns of A as vectors, returning a row vector of the sums of each column.
sum(Image); How could it be done with OpenCV?
Using cvReduce has worked for me. For example, if you need to store the column-wise sum of a matrix as a row matrix you could do this:
CvMat * MyMat = cvCreateMat(height, width, CV_64FC1);
// Fill in MyMat with some data...
CvMat * ColSum = cvCreateMat(1, MyMat->width, CV_64FC1);
cvReduce(MyMat, ColSum, 0, CV_REDUCE_SUM);
More information is available in the OpenCV documentation.
EDIT after 3 years:
The proper function for this is cv::reduce.
Reduces a matrix to a vector.
The function reduce reduces the matrix to a vector by treating the
matrix rows/columns as a set of 1D vectors and performing the
specified operation on the vectors until a single row/column is
obtained. For example, the function can be used to compute horizontal
and vertical projections of a raster image. In case of REDUCE_MAX and
REDUCE_MIN , the output image should have the same type as the source
one. In case of REDUCE_SUM and REDUCE_AVG , the output may have a
larger element bit-depth to preserve accuracy. And multi-channel
arrays are also supported in these two reduction modes.
OLD:
I've used ROI method: move ROI of height of the image and width 1 from left to right and calculate means.
Mat src = imread(filename, 0);
vector<int> graph( src.cols );
for (int c=0; c<src.cols-1; c++)
{
Mat roi = src( Rect( c,0,1,src.rows ) );
graph[c] = int(mean(roi)[0]);
}
Mat mgraph( 260, src.cols+10, CV_8UC3);
for (int c=0; c<src.cols-1; c++)
{
line( mgraph, Point(c+5,0), Point(c+5,graph[c]), Scalar(255,0,0), 1, CV_AA);
}
imshow("mgraph", mgraph);
imshow("source", src);
EDIT:
Just out of curiosity, I've tried resize to height 1 and the result was almost the same:
Mat test;
cv::resize(src,test,Size( src.cols,1 ));
Mat mgraph1( 260, src.cols+10, CV_8UC3);
for(int c=0; c<test.cols; c++)
{
graph[c] = test.at<uchar>(0,c);
}
for (int c=0; c<src.cols-1; c++)
{
line( mgraph1, Point(c+5,0), Point(c+5,graph[c]), Scalar(255,255,0), 1, CV_AA);
}
imshow("mgraph1", mgraph1);
cvSum respects ROI, so if you move a 1 px wide window over the whole image, you can calculate the sum of each column.
My c++ got a little rusty so I won't provide a code example, though the last time I did this I used OpenCVSharp and it worked fine. However, I'm not sure how efficient this method is.
My math skills are getting rusty too, but shouldn't it be possible to sum all elements in columns in a matrix by multiplying it by a vector of 1s?
For an 8 bit greyscale image, the following should work (I think).
It shouldn't be too hard to expand to different image types.
int imgStep = image->widthStep;
uchar* imageData = (uchar*)image->imageData;
uint result[image->width];
memset(result, 0, sizeof(uchar) * image->width);
for (int col = 0; col < image->width; col++) {
for (int row = 0; row < image->height; row++) {
result[col] += imageData[row * imgStep + col];
}
}
// your desired vector is in result