In case the title confused you. I want to remove the background around the object. The boundary is rather complex, so doing it by hand is time-consuming. However, I have several images of one object on different backgrounds.
So I've put these images on different layers, so the object on each layer is in the same place. Now I would like to combine all layers in one, so the object would persist, but different layers would be removed. Is there a function/filter/script that works this way? Taking pixels from different layers and if they are different removes them or makes them (more) transparent? While pixels that don't differ are left unchanged.
I've tried "addition" and "multiply" modes for layers, but they don't work that way - they still change pixels that are "the same".
With two images:
Set the top image to Difference
Get a new layer from the result: Layer>New from visible
Color-select the black with a low threshold.
Your selection is the pixels that are black, that are those where the difference between the images was 0, that are those that are identical in both images.
With more images
A solution likely uses a "median filter". Such a filter makes pixels "vote": a pixel is the most common values among the corresponding pixels in each of the source images. This is typically applied to remove random objects (tourists) in front of a fixed subject (building): take several shots, and the filter will keep the pixels from the building, removing the tourists.
There is a median filter in the GMIC plugin/filter suite. Otherwise if you have good computer skills (some install tweaks required) there is an experimental one in Python.
However the median filter doesn't erase the background so the technique is likely more complex than the tourist removal one. Can you show a sample picture?
Related
Perhaps my mind isn't mathematically competent enough to do this, but here it goes:
I am using Photoshop. I have 2 images taken from different heights. Both images have the same object in it (so the size of this object remains the same) but I am trying to resize both images so that this object is the same pixel size. That way I can properly measure the difference between other objects in the images with the proper ratio.
My end goal is to measure the differences of scars healing (before and after) using a same-size object in both images as a baseline.
To measure the difference in the photo, I have been counting pixels using the histogram feature:
Even though i changed the pixel width and height to roughly the same size, the 2 images have a drastically different number of pixels. So comparing the red or white from the before to the after won't make sense until I can get these to match.
Can anyone point me in the right direction here? How can I compare apples to apples here?
So went a different route here in case anyone was trying wondering what I did.
Rather than change the size of the images, just calculated the increase manually separately.
Let's say that I want to detect a character on a background image. The character can be tilted/moved in a number of ways, which will slightly change how we see him. Luckily, I have the sprite sheet for all of his possible positions. Is there a way to train tensorflow to detect objects/characters based on sprite sheets?
You can take different approaches:
1) I would first try out Template Matching. You slide your sprite over the image, and see if it matches the images. You do the same for the tilts, you tilt the sprite image, and slide the tilted sprite over the image. You do this for let's say every tenth of a degree, and take the best matching template.
2) If that's too computationally intensive, I would still use template matching, but only to gather data for a machine learning model. You can do template matching, then record the best match for a frame and the bounding boxes for that best match, and you can then use that to train an object detection network. There's more state-of-the-art stuff than this, but for ease-of-use I would Yolov3. it also has a tiny version, which is less accurate but way faster.
I found that a deep-learning-based method (e.g., 1) is much more robust than a non-deep-learning-based method (e.g., 2, using OpenCV).
https://www.remove.bg
How do I remove the background from this kind of image?
In the OpenCV example, Canny is used to detect the edges. But this step can be very sensitive to the image. The contour detection may end up with wrong contours. It is also difficult to determine which contours should be kept.
How a robust deep-learning method is implemented? Is any good example code? Thanks.
For that to work you need to use Unet. You can search for that on github.
Unet transofrm is: I->I.
Space of the image will become image (of same or similar size).
You need to have say 10.000 images with bg removed. People, (long hair people), cats, cars, shoes, T-shirts, etc.
So you set different backgrounds on all these images as source and prediction should be images with removed background.
You can also do a segmentation model and when you find the foreground you can remove the bg.
I have some problems with this assignment. Given an image of a street nameplate, like this one
I have to detect the nameplate and mark it on the image with a rectangle, obtaining something like this:
Nameplates can be rotated, scaled and in different lighting conditions. The procedure must be automatic.
What i have tried so far is to isolate the nameplate from the background. I've tried with different thresholding methods, but the problem is that i have different images and one single method doesn't work with all of them, due to different lighting condition and noise. What i've thought is to perform a pre-processing on the images, to reduce noise and normalize light, but, again, how to choose pre-processing steps that work with every image in my dataset? And what for images that don't need pre-processing?
Another problem is that there might be other signs in the image with writings on them and i have to ignore them. So i've thought i could isolate the nameplate by that blue outline, but i don't know if that can be done(or if it is convenient) with template matching, also considering that part of the outline could be cut off from the image.
So what i'm asking is: is there an automatic way to isolate/detect only that type of nameplates that have the blue outline on them, regardless of orientation, light conditions, shadows on them, noise in the image, etc? What steps would you follow?
Thank You
This question is very similar to that posed here.
My problem is that I have a map, something like this:
This map is made using 2D Perlin noise, and then running through the created heightmap assigning types and color values to each element in the terrain based on the height or the slope of the corresponding element, so pretty standard. The map array is two dimensional and the exact dimensions of the screen size (pixel-per-pixel), so at 1200 by 800 generation takes about 2 seconds on my rig.
Now zooming in on the highlighted rectangle:
Obviously with increased size comes lost detail. And herein lies the problem. I want to create additional detail on the fly, and then write it to disk as the player moves around (the player would simply be a dot restricted to movement along the grid). I see two approaches for doing this, and the first one that came to mind I quickly implemented:
This is a zoomed-in view of a new biased local terrain created from a sampled element of the old terrain, which is highlighted by the yellow grid space (to the left of center) in the previous image. However this system would require a great deal of modification, as, for example, if you move one unit left and up of the yellow grid space, onto the beach tile, the terrain changes completely:
So for that to work properly you'd need to do an excessive amount of, I guess the word would be interpolation, to create a smooth transition as the player moved the 40 or so grid-spaces in the local world required to reach the next tile over in the over world. That seems complicated and very inelegant.
The second approach would be to break up the grid of the original map into smaller bits, maybe dividing each square by 4? I haven't implemented this and I'm not sure how I would in a way that would actually increase detail, but I think that would probably end up being the best solution.
Any ideas on how I could approach this? Keep in mind it has to be local and on-the-fly. Just increasing the resolution of the map is something I want to avoid at all costs.
Rewrite your Perlin noise to be a function of position. Then you can increase the octaves (and thus the detail level) and resample the area at a higher resolution.