Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
This question does not appear to be about programming within the scope defined in the help center.
Closed 8 years ago.
Improve this question
I'm working on ways of collecting farm aerial images (images collected from a helicopter in a perpendicular fashion) that I'd want to stitch them together to build the whole photo of the area that's being covered and then I wanted to run analytics.
I'm assuming the images will come with [latitudue, longitude] coordinates, to help me determine the spots to place the images.
To understand the issues with this technology, I tried manually stitching pictures taken from my phone of some sample area in my back yard. I experienced that the edges don't usually look the same because they are being seen by the camera from different sides or angles. I guess this is a distortion in image that could potentially be fixed by ortho-rectification (not completely sure).
I quickly created the following picture to help explain my problem.
My question to you:
What are the algorithms/techniques used to do ortho-rectification?
What tools would best suit my needs: opencv, or processing or matlab or any other tool that could easily help in rectification of images and creating a mosaic photo?
What other issues should be considered in doing aerial imagery mosaic-ing and analytics?
Thank you!
Image stitching usually assumes that the camera center is fixed across all photos, and uses homographies to transform the images so that they seem continuous. When the fixed camera center assumption is not strictly valid, artifacts/distortions may appear due to the 3D of the scene. If the camera center moved by a small distance compared to the relief of the scene, "seamless image blending" techniques may be sufficient to blur out the distortions.
In more extreme cases, ortho-rectification is required. Ortho-rectification (Wikipedia entry) is the task of transforming an image observed from a given perspective camera into an orthographic (Wikipedia entry) and usually vertical point of view. The orthographic property is interesting because it makes the stitching of several images much easier. The following picture from Wikipedia is particularly clear (left is an orthographic or directional projection, right is a perspective or central projection):
The task of ortho-rectification usually requires having a 3D model of the scene, in order to map appropriately intensities observed by the perspective camera to their location with respect to the orthographic camera. In the context of aerial/satellite images, Digital Elevation Models (DEM) are often used for that purpose, but generally have the serious drawback of not including man-made structures (only Earth relief). The NASA provides freely the DEM acquired by the SRTM missions (DEM link).
Another approach, if you have two images acquired at different positions, you could try to do a 3D reconstruction using one of the stereo matching technique, and then to generate the ortho-rectified image by mapping the two images as seen by a third orthographic and vertical camera.
OpenCV has several interesting function for that purpose (e.g. stereo reconstruction, image mapping functions, etc) and might be more appropriate for intensive usage. Matlab probably has interesting functions as well, and might be more appropriate for quick tests.
First, rectification is some kind of warping, but not the one that you need. Regular rectification is used in stereo to ensure that matching points lie on the same row - not your case. Ortho-rectification warps perspective projection into orthographic - again not your case. Not only you lack a 3D model for this warping to calculate but also you dont need it since your perspective distortions are negligible and you already have image pretty close to ortho ( that is when the size of the objects is small compared to the viewing distance perspective effects are small).
You problems in aligning two images stem from small camera rotations between shots. To start fixing the problem you need to ensure that your images actually overlap by say 30%. To read about this see chapter 9 of this book.
What you need is to review a regular image stitching techniques that use homography to map two images. Note that doing so assumes that images are essentially flat. To find homography you can first manually select 4 points in one image and 4 matching points in another image and run openCV function findHomography(). Note that overlap is required to find the matches (in your picture there is no overlap). warpPerspective() can warp images for you after homogrphy is found.
"What are the algorithms/techniques used to do ortho-rectification?
If you want a good overview of the techniques, then the book "Multiple View Geometry in Computer Vision" by Hartley & Zisserman might be a good place to start: http://www.robots.ox.ac.uk:5000/~vgg/hzbook/
Andrew Zisserman also has some tutorials available here at www.robots.ox.ac.uk/~az/tutorials/ which might be more accessible/make it easier for you to find the particular technique you want to use.
"What tools would best suit my needs: opencv, or processing or matlab or any other tool that could easily help in rectification of images and creating a mosaic photo?"
OpenCV has a fair few number of tools available - take a look at Images stitching for starters. There's also a lot available for correcting distortion. However, it doesn't have to be the tool you use, there are others!
Related
I want to detect shape and then describe it (somehow) to compare it with server data.
So the first question is, is it possible to detect shape like blob with ARKit?
To be more specific, let's describe my usecase generally.
I want to scan image by phone, get the specific shape, send it on server, compare two images on server (server image is the real one, scanned image would be very similar) and then send back some data. I am not asking about server side, the only question about server side is what should I compare - images using OpenCV, some mathematical description of both images and try to find similarity, etc.).
If the question is hard to understand, let's split it on two easy questions:
1) How to scan 2D object by iPhone and save it (trim the specific shape from its background when object is black and background white).
2) Describe scanned object for comparision with almost the same object.
ARKit has no use here.
You will probably need a lot of CoreImage (for fixing perspective distortion and binarization) and OpenCV logic.
Perhaps Vision can help you a little bit with getting ROI from the entire frame, especially if the waveform image is located in some kind of rectangle.
Perhaps you can train a custom ML model that will recognize specific waveforms or waveforms in general to use with Vision.
In any case, it is not a trivial task.
If I have understand well, 3D 360 photos are created from a panorama photo, so I guess it should be possible to create a 3D photo (non 360) from a normal photo. But how? I did not find anything in Google! Any idea of what should I search??
So far, if nothing available (I don't think so), I'll try to duplicate the same photo in each eye. One of the pictures a little bit moved to the right, and the other one moved a little bit to the left. But I think the distortion algorithm is much more complicated.
Note: I'm also receiving answers here: https://plus.google.com/u/0/115463690952639951338/posts/4KdqFcqUTT9
I am in no way certain of this, but my intuition on how 3D 360 images are created in GoogleVR is this:
As you take a panorama image, it actually takes a series of images. As you turn the phone around, the perspective changes slightly with each image, not only by angle, but also offset (except in the unlikely event you spin the phone around its own axis). When it stitches together the final image, it creates one image for each eye, picking suitable images from the series so that it creates a 3D effect when viewed together. The same "area" of the image for each eye comes from a different source image.
You can't do anything similar with a single image. It's the multitude of images produced, each with a different perspective coming from the turning of the phone, that enables the algorithm to create a 3D image.
2D lacks a dimension hence cannot be converted to 3D just like that, but there are clever ways for example Google Pixel even though doesn't have 2 camera can make it seem like the image is 3D by applying some Machine learning algorithm that create the effect of perspective and depth by selective blurring.
3d photos can't be taken by normal but you can take 360 photos with normal camera ..... There are many apps via which you can do this ..... Also there are many algorithms to do it programmatically
I need some shape recognition advice. To be specific, I am data mining photos to recognize the presence of particular shapes, and these shapes may be connected together as well. i.e. the widget1 and widget2 I am interested in may be connected together by some frame.
These widgets may also be of different sizes, which may give issue with template matching techniques. For example, widget1 could be roughly 20x20 pixels in one picture and 100x100 pixels in another. Widget2 can be scaled differently in the same pictures. There can also be issues with pre-processing out some of the labeling/text that may be on these widgets as not to confuse any matching technique used.
Do you guys have any advice in which areas of image processing I should explore?
In summary, the issues are:
1) identifying known shapes
2) scaling differences can exist in the widgets between the photos
3) labeling on the widgets may exist that may confuse the algorithms used above - so should be pre-processed out
Thanks a bunch. If you guys can give some advice on suggested techniques and resources I should read-up on, that would be a great help!
This isn't necessarily a MATLAB or OpenCV question, but I would suggest having a look at MATLAB's Computer Vision toolbox and browse the Shape Descriptors sections of the OpenCV manual. Also, I realize this isn't a full answer, but there are many solutions to shape and object recognition, and the selected method depends heavily on the specifics of your problem.
I have read that it's possible to create a depth image from a stereo camera setup (where two cameras of identical focal length/aperture/other camera settings take photographs of an object from an angle).
Would it be possible to take two snapshots almost immediately after each other(on the iPhone for example) and use the differences between the two pictures to develop a depth image?
Small amounts of hand-movement and shaking will obviously rock the camera creating some angular displacement, and perhaps that displacement can be calculated by looking at the general angle of displacement of features detected in both photographs.
Another way to look at this problem is as structure-from-motion, a nice review of which can be found here.
Generally speaking, resolving spatial correspondence can also be factored as a temporal correspondence problem. If the scene doesn't change, then taking two images simultaneously from different viewpoints - as in stereo - is effectively the same as taking two images using the same camera but moved over time between the viewpoints.
I recently came upon a nice toy example of this in practice - implemented using OpenCV. The article includes some links to other, more robust, implementations.
For a deeper understanding I would recommend you get hold of an actual copy of Hartley and Zisserman's "Multiple View Geometry in Computer Vision" book.
You could probably come up with a very crude depth map from a "cha-cha" stereo image (as it's known in 3D photography circles) but it would be very crude at best.
Matching up the images is EXTREMELY CPU-intensive.
An iPhone is not a great device for doing the number-crunching. It's CPU isn't that fast, and memory bandwidth isn't great either.
Once Apple lets us use OpenCL on iOS you could write OpenCL code, which would help some.
I am building an application on night vision but i don't find any useful algorithm which I can apply on the dark images to make it clear. Anyone please suggest me some good algorithm.
Thanks in advance
With the size of the iphone lens and sensor, you are going to have a lot of noise no matter what you do. I would practice manipulating the image in Photoshop first, and you'll probably find that it is useful to select a white point out of a sample of the brighter pixels in the image and to use a curve. You'll probably also need to run an anti-noise filter and smoother. Edge detection or condensation may allow you to bold some areas of the image. As for specific algorithms to perform each of these filters there are a lot of Computer Science books and lists on the subject. Here is one list:
http://www.efg2.com/Lab/Library/ImageProcessing/Algorithms.htm
Many OpenGL implementations can be found if you find a standard name for an algorithm you need.
Real (useful) night vision typically uses an infrared light and an infrared-tuned camera. I think you're out of luck.
Of course using the iPhone 4's camera light could be considered "night vision" ...
Your real problem is the camera and not the algorithm.
You can apply algorithm to clarify images, but it won't make from dark to real like by magic ^^
But if you want to try some algorithms you should take a look at OpenCV (http://opencv.willowgarage.com/wiki/) there is some port like here http://ildan.blogspot.com/2008/07/creating-universal-static-opencv.html
I suppose there are two ways to refine the dark image. first is active which use infrared and other is passive which manipulates the pixel of the image....
The images will be noisy, but you can always try scaling up the pixel values (all of the components in RGB or just the luminance of HSV, either linear or applying some sort of curve, either globally or local to just the darker areas) and saturating them, and/or using a contrast edge enhancement filter algorithm.
If the camera and subject matter are sufficiently motionless (tripod, etc.) you could try summing each pixel over several image captures. Or you could do what some HDR apps do, and try aligning images before pixel processing across time.
I haven't seen any documentation on whether the iPhone's camera sensor has a wider wavelength gamut than the human eye.
I suggest conducting a simple test before trying to actually implement this:
Save a photo made in a dark room.
Open in GIMP (or a similar application).
Apply "Stretch HSV" algorithm (or equivalent).
Check if the resulting image quality is good enough.
This should give you an idea as to whether your camera is good enough to try it.