Well its a forum for posting the question in which you feel difficulty , well the same thing happen to me so i post the question here , i need to learn the code , understand it , what its doing and what we can do more with it
// Data for visual representation
int width = 512, height = 512;
Mat image = Mat::zeros(height, width, CV_8UC3);
// Set up training data
float labels[4] = {1.0, -1.0, -1.0, -1.0};
Mat labelsMat(3, 1, CV_32FC1, labels);
float trainingData[4][2] = { {501, 10}, {255, 10}, {501, 255}, {10, 501} };
Mat trainingDataMat(3, 2, CV_32FC1, trainingData);
// Set up SVM’s parameters
CvSVMParams params;
params.svm_type = CvSVM::C_SVC;
params.kernel_type = CvSVM::LINEAR;
params.term_crit = cvTermCriteria(CV_TERMCRIT_ITER, 100, 1e-6);
// Train the SVM
CvSVM SVM;
SVM.train(trainingDataMat, labelsMat, Mat(), Mat(), params);
Vec3b green(0,255,0), blue (255,0,0);
// Show the decision regions given by the SVM
for (int i = 0; i <2; ++i)
for (int j = 0; j <2; ++j)
{
Mat sampleMat = (Mat_<float>(1,2) << i,j);
float response = SVM.predict(sampleMat);
if (response == 1)
image.at<Vec3b>(j, i) = green;
else if (response == -1)
image.at<Vec3b>(j, i) = blue;
}
I know this code is for training data , but i want to know about its basic things , its basic understanding , which i think i didn't found on opencv documentation like why and when we use CV_8UC3 , and with which things this code is training
Thanks
image is an empty 3 channel matrix data, i.e. 512x512; R-G-B channels. At the end, this code draws the responses (predictions of SVM) onto that image - image at somewhere = green = (0,255,0). it is done in a for loop to create the lines from pointwise assigning.
the SVM model training is an internal process of this method, in which opencv uses a learning algorithm that can be found only looking at the source code. however, it is declared and described in the documentation that the parameters like svm_type, kernel_type, k_fold, grid, balanced, ... changes the behaviour of the method.
Related
I am new to OpenCV SVM. Is there a way to plot a graph or develop some visual content for the trained data set that is developed by OpenCV svm so that I can check if my training data is accurate and tune my SVM parameters accordingly ?
It is possible to visualize the SVM responses and can be done using OpenCV drawing capabilities.
This question is already old, but it shows up high in Google results for visualizing SVM. I answer for anyone who may find this question.
There is an SVM tutorial with example code here:
https://docs.opencv.org/2.4/doc/tutorials/ml/introduction_to_svm/introduction_to_svm.html
It contains a code for visualizing what the SVM has learned. Note that this example shows data on a plane so it will work only if your data is 2D.
So this code below (it comes from the linked OpenCV docs) creates a image 512 x 512 px and for each pixel checks if it was classified as belonging to one class or another (response 1 or -1). On top of it, the data points are marked so you are able to see if the SVM classification works well.
// Train the SVM
CvSVM SVM;
SVM.train(trainingDataMat, labelsMat, Mat(), Mat(), params);
Vec3b green(0,255,0), blue (255,0,0);
// Show the decision regions given by the SVM
for (int i = 0; i < image.rows; ++i)
for (int j = 0; j < image.cols; ++j)
{
Mat sampleMat = (Mat_<float>(1,2) << j,i);
float response = SVM.predict(sampleMat);
if (response == 1)
image.at<Vec3b>(i,j) = green;
else if (response == -1)
image.at<Vec3b>(i,j) = blue;
}
// Show the training data
int thickness = -1;
int lineType = 8;
circle( image, Point(501, 10), 5, Scalar( 0, 0, 0), thickness, lineType);
circle( image, Point(255, 10), 5, Scalar(255, 255, 255), thickness, lineType);
circle( image, Point(501, 255), 5, Scalar(255, 255, 255), thickness, lineType);
circle( image, Point( 10, 501), 5, Scalar(255, 255, 255), thickness, lineType);
This is a good approach if your data points are pairs of integers. If you have float data points you need to decide what resolution of visualization you want and make pixel represent not a 1 x 1 block but your demanded minimal resolution (e.g. 5.0e-27 x 2.0e-24)
I am having some issues with detecting specific "blobs" in a set of images. Not all images are the same, but I suppose the same parameters would be used to detect anyways.
If you zoom in, you will see small, yellow aphids on the leaf. My goal is to single these out and count them. I don't really need to do much to the image, just obtain a count of them.
Right now, I have this:
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using Emgu.CV;
using Emgu.CV.Features2D;
using Emgu.CV.Structure;
using Emgu.CV.Util;
namespace AphidCounter
{
class Program
{
static void Main(string[] args)
{
// Read image
Mat im_in = CvInvoke.Imread("myimage1.jpg", Emgu.CV.CvEnum.LoadImageType.Grayscale);
//Mat im_in = CvInvoke.Imread("myimage2.png", Emgu.CV.CvEnum.LoadImageType.Color);
Mat im = im_in;
CvInvoke.Threshold(im_in, im, 40, 255, Emgu.CV.CvEnum.ThresholdType.BinaryInv); // 60, 255, 1
//CvInvoke.NamedWindow("Blob Detector", Emgu.CV.CvEnum.NamedWindowType.AutoSize);
DetectBlobs(im, 0);
CvInvoke.WaitKey(0);
}
static void DetectBlobs(Mat im, int c)
{
int maxT = 50;
int minA = 125; // Minimum area in pixels
int maxA = 550; // Maximum area in pixels
SimpleBlobDetectorParams EMparams = new SimpleBlobDetectorParams();
SimpleBlobDetector detector;
EMparams.MinThreshold = 0;
EMparams.MaxThreshold = 100;
if (minA < 1) minA = 1;
EMparams.FilterByArea = true;
EMparams.MinArea = minA;
EMparams.MaxArea = maxA;
if (maxT < 1) maxT = 1;
EMparams.MinConvexity = (float)maxT / 1000.0F; // 0.67
EMparams.FilterByInertia = true;
EMparams.MinInertiaRatio = 0.01F;
EMparams.FilterByColor = true;
EMparams.blobColor = 0;
VectorOfKeyPoint keyPoints = new VectorOfKeyPoint();
detector = new SimpleBlobDetector(EMparams);
detector.DetectRaw(im, keyPoints);
Mat im_with_keypoints = new Mat();
Bgr color = new Bgr(0, 0, 255);
Features2DToolbox.DrawKeypoints(im, keyPoints, im_with_keypoints, color, Features2DToolbox.KeypointDrawType.DrawRichKeypoints);
// Show blobs
CvInvoke.Imwrite("keypoints1.jpg", im_with_keypoints);
CvInvoke.Imshow("Blob Detector " + keyPoints.Size, im_with_keypoints);
System.Console.WriteLine("Number of keypoints: " + keyPoints.Size);
}
}
}
However, this is the result:
Am I not getting the parameters right? Or is there something else that I'm missing?
It is not because of some wrong parameters. The image segmentation part itself has its limitation.
Grayscale based thresholding may not work when the contrast between the blob and the background is very low. Yet a threshold value around 160 is quite tolerable in this example but not any accurate.
I would suggest to go for colour based thresholding since there is a decent colour gap.
Here is a C++ implementation of colour based thresholding. Blobs are filtered using the same SimpleBlobDetector.
I have converted the image from RGB to ‘Lab’ for better segmentation.
As the image provided is too huge, it took more time to process. So I cropped a key part of the image and tuned the blob params for the same. So I provide the cropped image too (755 x 494px).
Colour based thresholding and blob filtering:
#include "opencv2\imgproc\imgproc.hpp";
#include "opencv2\highgui\highgui.hpp";
#include "opencv2\features2d\features2d.hpp";
using namespace cv;
using namespace std;
void main()
{
char image_path[] = "E:/Coding/media/images/leaf_small.jpg";
Mat img_color, img_lab, img_thresh, img_open, img_close, img_keypoints;
img_color = imread(image_path, IMREAD_ANYCOLOR);
//Convert image to CIE Lab colorspace for better colour based segmentation
cvtColor(img_color, img_lab, CV_BGR2Lab);
//create window before creating trackbar
namedWindow("win_thresh", WINDOW_NORMAL);
namedWindow("win_blob", WINDOW_NORMAL);
//Using trackbar calculate the range of L,a,b values to seperate blobs
int low_L = 150, low_A = 0, low_B = 155,
high_L = 255, high_A = 255, high_B = 255;
//*Use trackbars to caliberate colour thresholding
createTrackbar("low_L", "win_thresh", &low_L, 255);
createTrackbar("low_A", "win_thresh", &low_A, 255);
createTrackbar("low_B", "win_thresh", &low_B, 255);
createTrackbar("high_L", "win_thresh", &high_L, 255);
createTrackbar("high_A", "win_thresh", &high_A, 255);
createTrackbar("high_B", "win_thresh", &high_B, 255);
int minArea = 35, maxArea = 172, minCircularity = 58, minConvexity = 87, minInertiaRatio = 21;
//Use trackbar and set Blob detector parameters
createTrackbar("minArea", "win_blob", &minArea, 200);
createTrackbar("maxArea", "win_blob", &maxArea, 200);
createTrackbar("minCircular", "win_blob", &minCircularity, 99);
createTrackbar("minConvex", "win_blob", &minConvexity, 99);
createTrackbar("minInertia", "win_blob", &minInertiaRatio, 99);
SimpleBlobDetector::Params params;
vector<KeyPoint> keypoints;
while (waitKey(1) != 27) //press 'esc' to quit
{
//inRange thresholds basedon the Scalar boundaries provided
inRange(img_lab, Scalar(low_L, low_A, low_B), Scalar(high_L, high_A, high_B), img_thresh);
//Morphological filling
Mat strucElement = getStructuringElement(CV_SHAPE_ELLIPSE, Size(5, 5), Point(2, 2));
morphologyEx(img_thresh, img_close, MORPH_CLOSE, strucElement);
imshow("win_thresh", img_close);
//**SimpleBlobDetector works only in inverted binary images
//i.e.blobs should be in black and background in white.
bitwise_not(img_close, img_close); // inverts matrix
//Code crashes if minArea or any miin value is set to zero
//since trackbar starts from 0, it is adjusted here by adding 1
params.filterByArea = true;
params.minArea = minArea + 1;
params.maxArea = maxArea + 1;
params.filterByCircularity = true;
params.filterByConvexity = true;
params.filterByInertia = true;
params.minCircularity = (minCircularity + 1) / 100.0;
params.minConvexity = (minConvexity + 1) / 100.0;
params.minInertiaRatio = (minInertiaRatio + 1) / 100.0;
SimpleBlobDetector detector(params);
detector.detect(img_close, keypoints);
drawKeypoints(img_color, keypoints, img_keypoints, Scalar(0, 0, 255), DrawMatchesFlags::DEFAULT);
stringstream displayText;
displayText = stringstream();
displayText << "Blob_count: " << keypoints.size();
putText(img_keypoints, displayText.str(), Point(0, 50), CV_FONT_HERSHEY_PLAIN, 2, Scalar(0, 0, 255), 2);
imshow("win_blob", img_keypoints);
}
return;
}
Output Screenshot
Tune the blob parameters according to the actual HD image.
Since the veins of the leaf are almost of the same colour and intensity of the aphid, this method also may utterly fail when an aphid sits close to or exactly on top of a vein.
This can be an ad-hoc fix but not robust enough.
There got to be a simple and robust method to achieve the result, using some filters, transformation or edge detection. Please share any other optimal solution if available.
EDIT: Opting Grayscale thresholding as previous approach failed
Colour thresholding approach failed for this_image
Colour based thresholding has a very narrow bandwidth, if the image falls within the bandwidth the accuracy will be really good, on the other hand colour shifts totally ruin the accuracy.
Since you will be processing 100s of images, colour thresholding may not be suitable.
I tried normal Grayscale thresholding with some morphological erosion and filling, and got a decent accuracy. Also Grayscale thresholding has better immunity to colour shifts.
Additionally we have auto thrsholding option using OTSU Thresholding which selects the threshold value based on the image.
Code snippet:
threshold(img_gray, img_thresh, 0, 255, THRESH_OTSU);
Mat strucElement = getStructuringElement(CV_SHAPE_ELLIPSE, Size(3, 3), Point(1, 1));
morphologyEx(img_thresh, img_open, MORPH_OPEN, strucElement);
Rest of the code remains the same.
Parameter values:
minArea = 75, maxArea = 1000, minCircularity = 50, minConvexity = 20, minInertiaRatio = 15
The white ants are hard to differentiate from aphids as we are not using colour information. So the min_area has to be carefully tuned in order to exclude them.
Processed images can be found here img_1, img_2.
Tweak the morphology methods and blob parameters to obtain an optimal average count.
I am struggling with finding the appropriate contour algorithm for a low quality image. The example image shows a rock scene:
What I am trying to achieve is to find contours arround features such as:
light areas
dark areas
grey1 areas
grey2 areas
etc. until grey-n areas
(The number of areas shall be a parameter of choice)
I do not want to take a simple binary-threshold but rather use some sort of contour-finding (for example watershed or other). The major feature-lines shall be kept, noise within a feature-are can be flattened.
The result of my code can be seen on the images to the right.
Unfortunately, as you can easily tell, the colors do not really represent the original large-scale image features! For example: check out the two areas that I circled with red - these features are almost completely flooded with another color. What I imagine is that at least the very light and the very dark areas are covered by its own color.
cv::Mat cv_src = cv::imread(argv[1]);
cv::Mat output;
cv::Mat cv_src_gray;
cv::cvtColor(cv_src, cv_src_gray, cv::COLOR_RGB2GRAY);
double clipLimit = 0.1;
cv::Size titleGridSize = cv::Size(8,8);
cv::Ptr<cv::CLAHE> clahe = cv::createCLAHE(clipLimit, titleGridSize);
clahe->apply(cv_src_gray, output);
cv::equalizeHist(output, output);
cv::cvtColor(output, cv_src, cv::COLOR_GRAY2RGB);
// Create binary image from source image
cv::Mat bw;
cv::cvtColor(cv_src, bw, cv::COLOR_BGR2GRAY);
cv::threshold(bw, bw, 180, 255, cv::THRESH_BINARY);
// Perform the distance transform algorithm
cv::Mat dist;
cv::distanceTransform(bw, dist, cv::DIST_L2, CV_32F);
// Normalize the distance image for range = {0.0, 1.0}
cv::normalize(dist, dist, 0, 1., cv::NORM_MINMAX);
// Threshold to obtain the peaks
cv::threshold(dist, dist, .2, 1., cv::THRESH_BINARY);
// Create the CV_8U version of the distance image
cv::Mat dist_8u;
dist.convertTo(dist_8u, CV_8U);
// Find total markers
std::vector<std::vector<cv::Point> > contours;
cv::findContours(dist_8u, contours, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_SIMPLE);
int ncomp = contours.size();
// Create the marker image for the watershed algorithm
cv::Mat markers = cv::Mat::zeros(dist.size(), CV_32S);
// Draw the foreground markers
for (int i = 0; i < ncomp; i++)
cv::drawContours(markers, contours, i, cv::Scalar::all(i+1), -1);
// Draw the background marker
cv::circle(markers, cv::Point(5,5), 3, CV_RGB(255,255,255), -1);
// Perform the watershed algorithm
cv::watershed(cv_src, markers);
// Generate random colors
std::vector<cv::Vec3b> colors;
for (int i = 0; i < ncomp; i++)
{
int b = cv::theRNG().uniform(0, 255);
int g = cv::theRNG().uniform(0, 255);
int r = cv::theRNG().uniform(0, 255);
colors.push_back(cv::Vec3b((uchar)b, (uchar)g, (uchar)r));
}
// Create the result image
cv::Mat dst = cv::Mat::zeros(markers.size(), CV_8UC3);
// Fill labeled objects with random colors
for (int i = 0; i < markers.rows; i++)
{
for (int j = 0; j < markers.cols; j++)
{
int index = markers.at<int>(i,j);
if (index > 0 && index <= ncomp)
dst.at<cv::Vec3b>(i,j) = colors[index-1];
else
dst.at<cv::Vec3b>(i,j) = cv::Vec3b(0,0,0);
}
}
// Show me what you got
imshow("final_result", dst);
I think you can use a simple clustering such as k-means for this, then examine the cluster centers (or the mean and standard deviations of each cluster). I quickly tried it in matlab.
im = imread('tvBqt.jpg');
gr = rgb2gray(im);
x = double(gr(:));
idx = kmeans(x, 4);
cl = reshape(idx, 600, 472);
figure,
subplot(1, 2, 1), imshow(gr, []), title('original')
subplot(1, 2, 2), imshow(label2rgb(cl), []), title('clustered')
The result:
You could try using SLIC Superpixels. I tried it and showed some good results. You could vary the parameters to get better clustering.
SLIC Superpixels
SLIC Superpixels with OpenCV C++
SLIC Superpixels with OpenCV Python
I have optical flow stored in a 2-channel 32F matrix. I want to visualize the contents, what's the easiest way to do this?
How do I convert a CV_32FC2 to RGB with an empty blue channel, something imshow can handle? I am using OpenCV 2 C++ API.
Super Bonus Points
Ideally I would get the angle of flow in hue and the magnitude in brightness (with saturation at a constant 100%).
imshow can handle only 1-channel gray-scale and 3-4 channel BRG/BGRA images. So you need do a conversion yourself.
I think you can do something similar to:
//extraxt x and y channels
cv::Mat xy[2]; //X,Y
cv::split(flow, xy);
//calculate angle and magnitude
cv::Mat magnitude, angle;
cv::cartToPolar(xy[0], xy[1], magnitude, angle, true);
//translate magnitude to range [0;1]
double mag_max;
cv::minMaxLoc(magnitude, 0, &mag_max);
magnitude.convertTo(magnitude, -1, 1.0 / mag_max);
//build hsv image
cv::Mat _hsv[3], hsv;
_hsv[0] = angle;
_hsv[1] = cv::Mat::ones(angle.size(), CV_32F);
_hsv[2] = magnitude;
cv::merge(_hsv, 3, hsv);
//convert to BGR and show
cv::Mat bgr;//CV_32FC3 matrix
cv::cvtColor(hsv, bgr, cv::COLOR_HSV2BGR);
cv::imshow("optical flow", bgr);
cv::waitKey(0);
The MPI Sintel Dataset provides C and MatLab code for visualizing computed flow. Download the ground truth optical flow of the training set from here. The archive contains a folder flow_code containing the mentioned source code.
You can port the code to OpenCV, however, I wrote a simple OpenCV wrapper to easily use the provided code. Note that the method MotionToColor is taken from the color_flow.cpp file. Note the comments in the listing below.
// Important to include this before flowIO.h!
#include "imageLib.h"
#include "flowIO.h"
#include "colorcode.h"
// I moved the MotionToColor method in a separate header file.
#include "motiontocolor.h"
cv::Mat flow;
// Compute optical flow (e.g. using OpenCV); result should be
// 2-channel float matrix.
assert(flow.channels() == 2);
// assert(flow.type() == CV_32F);
int rows = flow.rows;
int cols = flow.cols;
CFloatImage cFlow(cols, rows, 2);
// Convert flow to CFLoatImage:
for (int i = 0; i < rows; i++) {
for (int j = 0; j < cols; j++) {
cFlow.Pixel(j, i, 0) = flow.at<cv::Vec2f>(i, j)[0];
cFlow.Pixel(j, i, 1) = flow.at<cv::Vec2f>(i, j)[1];
}
}
CByteImage cImage;
MotionToColor(cFlow, cImage, max);
cv::Mat image(rows, cols, CV_8UC3, cv::Scalar(0, 0, 0));
// Compute back to cv::Mat with 3 channels in BGR:
for (int i = 0; i < rows; i++) {
for (int j = 0; j < cols; j++) {
image.at<cv::Vec3b>(i, j)[0] = cImage.Pixel(j, i, 0);
image.at<cv::Vec3b>(i, j)[1] = cImage.Pixel(j, i, 1);
image.at<cv::Vec3b>(i, j)[2] = cImage.Pixel(j, i, 2);
}
}
// Display or output the image ...
Below is the result when using the Optical Flow code and example images provided by Ce Liu.
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