charles1208's profile - activity

cvtColor vs reading image directly I have the following: string input = "test.jpg"; cv::Mat tmp = cv::imread(input, CV

cvtColor vs reading image directly I have the following: string input = "test.jpg"; cv::Mat tmp = cv::imread(input, CV

I have the following code:

string src = "a.jpg";
string dest = "a_out.jpg";

cv::Mat im = cv::imread(src, CV_LOAD_IMAGE_COLOR);
cv::imwrite(dest, im);
cv::Mat im2 = cv::imread(dest, CV_LOAD_IMAGE_COLOR);
isMatEqual(im,im2);

where:

bool isMatEqual(const cv::Mat & a, const cv::Mat & b)
{
    if ( (a.rows != b.rows) || (a.cols != b.cols) )
        return false;
    cv::Scalar s = sum( a - b );
    return (s[0]==0) && (s[1]==0) && (s[2]==0);
}

However, I get a return that is false. Is there something wrong my reading/writing of the images?

imread, imwrite does not give similar image I have the following code: string src = "a.jpg"; string dest = "a_out.jpg";

Calculating original image points with perspectiveTransform() Given a pair of images, imgA and imgB, i resized them firs

I have an image that I want to stretch to cover the entire screen after transformation.

My code is as such:

src = np.array([[0, 0], [997, 102], [1000, 600], [0, 995]], np.float32)
dst = np.array([[0, 0], [997, 0], [1000, 995], [0, 995]], np.float32)
M = cv2.getPerspectiveTransform(src, dst)
warped = cv2.warpPerspective(img, M, (1000, 1000))

where the src coordinates refer to the 4 red points in the image, in clockwise order.

The dst coordinates refer to the 2 red points on the left, and the 2 blue points on the right of the image, in clockwise order.

The idea is that I want to "stretch" the image such that the 2 red points on the right correspond to the 2 blue points.

However, the warped image turns out to be:

Is there something I did wrong, or assumed wrongly about the usage of the functions?

Using getPerspective to warp image to forward facing I have an image that I want to stretch to cover the entire screen a

@berak unfortunately i do not know the distance to the camera; i only know the actual distance between the 2 blue lines.

Finding the real-world distance of object from pixel coordinates I have a picture of a supermarket shelf, with its top-m

Finding the real-world distance of object from pixel coordinates I have a picture of a supermarket shelf, with its top-m

There is a change to the scope of my question. I will delete this current one and rephrase it better with another one. T

Eliminating erroneous non-converging lines I am trying to detect shelf lines at a supermarket. In the picture attach

Not exactly. What i am looking for is more of like to simulate how an image of an object would look like under different

Recreating lighting based on sample image Hi, I am wondering if there is any current research/implementation to this pr

Recreating lighting based on sample image Hi, I am wondering if there is any current research/implementation to this pr

Recreating lighting based on sample image Hi, I am wondering if there is any current research/implementation to this pr

Given an image of many items, with all of its bounding box known in pixel coordinates.

I am trying to extract a region (surrounding) around each of the items, calculate its descriptors and features using AKAZE, to do comparison with one another.

However I realised that this might be too slow, since it involves:

1) cropping each and every single item to generate many images then,

2) detecting and computing on each image to generate the keypoints and descriptors.

Alternatively, to speed things up, I was thinking of:

1) Resizing the entire image, then perform the detecting and computing of keypoints once.

2) Then to obtain the keypoint of a particular object, we simply retrieve the set of precalculated keypoints corresponding to the objects location.

My question is this method functionally sound, and that if there are any consequences to this?

Cropping keypoints vs. cropping image and finding keypoints Given an image of many items, with all of its bounding box k

Not really. For this it ran okay under release, but under debug, the following error appears: Assertion failed in cv::Tl

knnMatch gives different results between debug and release Mode I am trying to use the BFMatcher class for Akaze auto b

knnMatch gives different results between debug and release Mode I am trying to use the BFMatcher class for Akaze auto b

Yes I am using windows/VS, but i checked that the debug are linked to the *d.lib files while the release are linked to t

knnMatch gives different results between debug and release Mode I am trying to use the BFMatcher class for Akaze auto b

I got a strange error that I couldn't figure out where i went wrong.. I am trying to port from Python over to C++ the following cropping code:

In Python:

# x, y, w, h refers to the ROI for which the image is to be cropped.
img = cv2.imread('test.jpg')
cropped_img = img[y: y + h, x: x + w]

In C++:

auto img = cv::imread("test.jpg");
auto resized_img = img(cv::Rect(x, y, w, h));

The values used are exactly the same. However the c++ version throws the exception:

OpenCV Error: Assertion failed (0 <= roi.x && 0 <= roi.width && roi.x + roi.width <= m.cols && 0 <= roi.y && 0 <= roi.height && roi.y + roi.height <= m.rows) in cv::Mat::Mat

Cropping of images in c++ vs python I got a strange error that I couldn't figure out where i went wrong.. I am trying to