BhanuKiran's profile - activity

Yes, you are right. I didn't realize that i need to convert it to float32/float64. I just used object = np.transpose(n

Yes, you are right. I didn't realize that i need to convert it to float32/float64. I just used object = np.transpose(np

Perform a Principle component analysis(PCA) I am trying to perform PCA on a binary mask of an image. Mask is of shape(48

Image restoration Hi, I am looking to restore images which are taken from a moving platform. Restoration filter woul

I am sorry. Now I am realizing that my question does make any sense. If i cut the image along the line as left image and

Split an image into 2 parts based on a line Hi, I am looking to split an image into two parts. After doing some math

I am sorry. Now I am realizing that my question does make any sense. If i cut the image along the line as left image and

What do u mean by overlapping ? From HoughLinesP function i get two points on the line. I extrapolated the line to get t

Split an image into 2 parts based on a line Hi, I am looking to split an image into two parts. After doing some math

Split an image into 2 parts based on a line Hi, I am looking split an image into two parts. After doing some math in

Fit a rectangle/any specific structure in an image Hi, I am trying to identify the bins in the image and separate them

Obtain continuous closed contour Hi, This contour lines are directly produced from the code taken from "Finding contou

Separate a closed contour from an image Hi, Above is the attached image of the lobster after applying the canny edge.

Separate a closed contour from an image Hi, Above is the attached image of the lobster after applying the canny edge.

Thank you for the comment. Previously, i was working with only 14 image pair. Now i tried with 29. I see the epipolar lines matching nicely unlike above image. But RMS error = 1.56 average epipolar error = 2.6 . still remain higher.

cv::CommandLineParser parser(argc, argv, "{w|9|}{h|6|}{s|20|}{nr||}{help||}{@input|stereo_calib.xml|}");
float squareSize = parser.get<float>("s");

My square size is 20mm. The argument for variable S should in mm right.

I ran stereo calibration

RMS error = 2.5624 average epipolar err = 2.79867

I am not sure, if these values make sense.

Left Image

(/upfiles/15011100969575928.jpg)

Right Image

Rectified image

I am using windows. There is no error after compiling the code but no windows is showing up and code is terminating instantly.

Its working well with point grey software. I also interfaced with opencv C++ following their tutorial. Actually, i want to input video for camera calibration.

Hi , Thanks for response. I was doing the way you mentioned before. But for the purpose of calibration using opencv, i want to choose video as an input. In the file "in_VID5.xml"

<Input>"1"</Input>

is not working. I am wondering if there is way i can input point grey camera video as an input for the camera calibration.

I have a point grey camera USB3 type and connect it to my computer using USB 3.0 port.

I came across this code recently. But not sure how to use.

#include "opencv2\opencv.hpp"
#include <stdint.h>

using namespace cv;
using namespace std;

const int fps = 20;

int main(int argv, char** argc) {
    Mat frame;
    VideoCapture vid(0);

if (!vid.isOpened()) {
    return -1;
}

while (vid.read(frame)) {
    imshow("Webcam", frame);

    if (waitKey(1000 / fps) >= 0)
        break;
    }
    return 1;
}

I am not sure how to use this code with my camera. Please help.

I am sorry, my question isn't clear I guess. I edited it.

I am trying to take video/lot of pictures of environment and convert it into 3D model

Hi

I am thinking of model an outdoor environment (probable around 5000 sq.ft or even more). Later on,i want to estimate camera pose based on the image that i captured in the environment.(by matching my descriptors of the model with that in the image).

I saw a tutorial in OpenCV - Real time pose estimation of a textured object. But is mentioned that the program only works for a planar objects.

I am not sure which software should i need to use to make a 3D textured model from environment. Thinking of flying quad copter over a small area and converting it into model.

Any thoughts about this or web links that can address this, would be greatly appreciated.

Thanks

Thanks, I didn't know that!!

I am using Visual Studio, OpenCV 3.2.0.

I want to draw a line on an image . I know the coefficients (a,b,c values in ax+by+c=0). Is there a way to draw a line rather than specifying two points?

Thanks for your reply. It's working

Hi I have 2 vectors and i want to perform cross and dot product operations on them.

double va[] = {1.0, 2.0, 3.0};
double vb[] = {0.0, 0.0, 1.0};
double vc[3];

cv::Mat VA(1, 3, CV_32FC1, va);
cv::Mat VB(1, 3, CV_32FC1, vb);

std::cout << VA << std::endl;

When i print them my both va and VA are not the same !!!!

Hi, It looked like this

P2

820 615

255

188 188 188 190 .........

Hi

I am using C++ openCV-3.2.0 version. I read .pgm image

cv::Mat pgm_image;

pgm_image = cv::imread("image.pgm");

and i am trying to access elements of pgm_image (cv::Mat)

i tried using std::cout << pgm_image .at<double>(0,0) << std::endl;

I am unable to print anything and program is aborting ? Am i doing anything wrong here ?

Code is printing the matrix when i use std::cout << pgm_image << std::endl;

Hi,

I am using C++ openCV-3.2.0 version.

I am also using a library for line segment detection,which takes only .pgm image of data type double *. I used

cv::Mat pgm_image;

pgm_image = cv::imread("image.pgm");

double *I = pgm_image.ptr<double>(0);

to read the image and to for getting pointer. And i tried parsing double * into the library that i am using. Is there any way around this? or Is there a way read .pgm file directly as double* ?