What is the convention for image dimensions in Torch? I'm not a torch user, but am reading some code and I need to know if Torch typically represents an RGB image as {height, width, channels} or as {channels, width, height} as I think I'm reading out of the code I'm looking at.
If you are using the image package (which is the usual way to go), then it usually represents images as tensors of size (channels, height, width).
Related
I am new to the field of medical imaging - and trying to solve this (potentially basic problem). For a machine learning purpose, I am trying to standardize and normalize a library of DICOM images, to ensure that all images have the same rotation and are at the same scale (e.g. in mm). I have been playing around with the Mango viewer, and understand that one can create transformation matrices that might be helpful in this regard. I have however the following basic questions:
I would have thought that a scaling of the image would have changed the pixel spacing in the image header. Does this tag not provide the distance between pixels, and should this not change as a result of scaling?
What is the easiest way to standardize a library of images (ideally in python)? Is it possible and should one extract a mean pixel spacing across all images, and then scaling all images to match that mean? or is there a smarter way to ensure consistency in scaling and rotation?
Many thanks in advance, W
Does this tag not provide the distance between pixels, and should this
not change as a result of scaling?
Think of the image voxels as fixed units of space, which are sampling your image. When you apply your transform, you are translating/rotating/scaling your image around within these fixed units of space. That is, the size and shape of the voxels doesn't change. They just sample different parts of your image.
You can resample your image by making your voxels bigger or smaller or changing their shape (pixel spacing), but this can be independent of the transform you are applying to the image.
What is the easiest way to standardize a library of images (ideally in
python)?
One option is FSL-FLIRT, although it only accepts data in NIFTI format, so you'd have to convert your DICOMs to NIFTI. There is also this Python interface to FSL.
Is it possible and should one extract a mean pixel spacing across all
images, and then scaling all images to match that mean? or is there a
smarter way to ensure consistency in scaling and rotation?
I think you'd just to have pick a reference image to register all your other images too. There's no right answer: picking the highest resolution image/voxel dimensions or an average or some resampling into some other set of dimensions all sound reasonable.
I'm scaling a QImage, currently as so (I understand there may be more elegant ways):
img.setDotsPerMeterX(img.dotsPerMeterX() * 2);
img.setDotsPerMeterY(img.dotsPerMeterY() * 2);
When I save:
img.save("c:\\users\\me\\desktop\\test.jpg");
and subsequently open and print the image from Photoshop, it is, as expected, half of the physical size of the same image without the scaling applied.
However, when I simply print the scaled QImage, directly from code:
myQPainter.drawImage(0,0,img);
the image prints at the original physical size - not scaled to half the physical size.
I'm using the same printer in each case; and, as far as I can tell, the settings are consistent between both print cases.
Am I misunderstanding something? The end goal is to successfully scale and print the scaled image directly from code.
If we look at the documentation for setDotsPerMeterX it states: -
Together with dotsPerMeterY(), this number defines the intended scale and aspect ratio of the image, and determines the scale at which QPainter will draw graphics on the image. It does not change the scale or aspect ratio of the image when it is rendered on other paint devices.
I expect that the reason for the latter case being the original size is that the image has already been drawn before the call to the functions to set the dots per meter. Or alternatively, set the dots per meter on the original image, before loading its content.
In contrast, when saving, it appears that the device which you save to is copying the values you have set for the dots per meter on the image, then drawing to that device.
I would expect creating a second QImage, setting its dots per meter, then copying from the original to that second image, it would achieve the result you're looking for. Alternatively, you may just be able to set the dots per meter before loading the content on the original QImage.
I create a big image stitched out of many single microscope images.
Suddenly, (after several month of working properly) the stitched overview images became blurry and they are containing strange structural artefacts like askew lines (not the rectangulars, they are because of not perfect stitching)
If I open any particular tile in full size, they are not blurry and the artefacts are hardly observable. (Consider, the image below is already 4x scaled)
The overview image is created manually by scaling each tile using QImage::scaled and copying all of them to the corresponding region in the big image. I'm not using opencv's stitching.
I assume, this happens because of image contents, because most if the overview images are ok.
The question is, how can I avoid such hardly observable artefacts to become very clearly visible after scaling? Is there some means in OpenCV or QImage?
Is there any algorithms to find out, if image content could lead to such effect for defined scale-factor?
Many thanks in advance!
Are you sure the camera is calibrated properly? That the lightning is uniform? Is the lens clear? Do you have electrical components that interfere with the camera connection?
If you add image frames of photos on a uniform material (or non-uniform material, moved randomly for significant time), the resultant integrated image should be completely uniform.
If your produced image is not uniform, especially if you get systematic noise (like the apparent sinusoidal noise in the provided pictures), write a calibration function that transforms image -> calibrated image.
Filtering in Fourier space is another way to filter out the noise but considering that the image is rotated you will lose precision, and you'll be cutting off components of the real signal, too. The following empiric method will reduce the noise in your particular case significantly:
ground_output: composite image with per-pixel sum of >10 frames (more is better) over uniform material (e.g. excited slab of phosphorus)
ground_input: the average(or sqrt(sum of px^2)) in ground_output
calib_image: ground_input /(per px) ground_output. Saved for the session, or persistent in a file (important: ensure no lossy compression! (jpeg)).
work_input: the images to work on
work_output = work_input *(per px) calib_image: images calibrated for systematic noise.
If you can't create a perfect ground_input target such as having a uniform material on hand, do not worry too much. If you move any material uniformly (or randomly) for enough time, it will act as a uniform material in this case (think of a blurred photo).
This method has the added advantage of calibrating solitary faulty pixels that ccd cameras have (eg NormalPixel.value(signal)).
If you want to have more fun you can always fit the calibration function to something more complex than a zero-intercept line (steps 3. and 5.).
I suggest scaling the image with some other software to verify if the artifacts are in fact caused by Qt or are inherent in the image you've captured.
The askew lines look a lot like analog tv interference, or CCTV noise induced by 50 or 60 Hz power lines running alongside the signal cable or some other electrical interference on the signal.
If the image distortion is caused by signal interference then you can try to mitigate it by moving the signal lines away from whatever could be the source of the problem, or fit something to try to filter the noise (baluns for example).
Can I set up an opencl image so that coordinate access past the boundary
of the image will return the mirror image?
For example, if image is of dimensions width by height, then
read_floati(width, 0)
will return
read_floati(width-2,0)
Yes. Read section 6.11.13 of the OpenCL specification. OpenCL images are read using (for example) read_imagef function and that function takes a sampler which can be set up for mirroring using CLK_ADDRESS_MIRRORED_REPEAT.
I have a QImage of size 12x12 in GIF format. I want to rotate it on certain angle with very high frequency. My application involves a robot so when it changes its orientation(which it does very frequently) my QImage in simulation should also be rotated but it causes loss of information. I am doing it something like below.
robot_transform.rotate(angle);
*robot2 = robot->transformed(robot_transform,Qt::SmoothTransformation);
*robot2= robot2->scaled(12,12, Qt::KeepAspectRatio,Qt::SmoothTransformation);
I need suggestions that whats wrong in this approach and secondly is there any other optimal approach for the desired application?
Thanks
I would increase the resolution of the source image to at least double. Rotating an image to non-90-degree angles will cause loss of pixel information. An higher res source can compensate for that.
Most sprite based animations use pre-rendered images for each possible angle.
The problem is the scaling afterwards, you need to crop the center of the image. You can do this with QImage::copy.