Closed. This question needs details or clarity. It is not currently accepting answers.
Want to improve this question? Add details and clarify the problem by editing this post.
Closed 8 years ago.
Improve this question
I was given this question on a job interview and think I really messed up. I was wondering how others would go about it so I could learn from this experience.
You have one image from a surveillance video located at an airport which includes line of people waiting for check-in. You have to assess if the line is big/crowded and therefore additional clerks are necessary. You can assume anything that may help your answer. What would you do?
I told them I would try to
segment the area containing people from the rest by edge detection
use assumptions on body contour such as relative height/width to denoise unwanted edges
use color knowledges; but then they asked how to do that and I didn't know
You failed to mention one of the things that makes it easy to identify people standing in a queue — the fact that they aren't going anywhere (at least, not very quickly). I'd do it something like this (Warning: contains lousy Blender graphics):
You said I could assume anything, so I'll assume that the airport's floor is a nice uniform green colour. Let's take a snapshot of the queue every 10 seconds:
We can use a colour range filter to identify the areas of floor that are empty in each image:
Then by calculating the maximum pixel values in each of these images, we can eliminate people who are just milling around and not part of the queue. Calculating the queue length from this image should be very easy:
There are several ways of improving on this. For example, green might not be a good choice of colour in Dublin airport on St Patrick's day. Chequered tiles would be a little more difficult to segregate from foreground objects, but the results would be more reliable. Using an infrared camera to detect heat patterns is another alternative.
But the general approach should be fairly robust. There's absolutely no need to try and identify the outlines of individual people — this is really very difficult when people are standing close together.
I would just use a person detector, for example OpenCV's HOG people detection:
http://docs.opencv.org/modules/gpu/doc/object_detection.html
or latent svm with the person model:
http://docs.opencv.org/modules/objdetect/doc/latent_svm.html
I would count the number of people in the queue...
I would estimate the color of the empty floor, and go to a normalized color space (like { R/(R+G+B), G/(R+G+B) } ). Also do this for the image you want to check, and compare these two.
My assumption: where the difference is larger than a threshold T it is due to a person.
When this is happening for too much space it is crowded and you need more clerks for check-in.
This processing will be way more robust than trying to recognize and count individual persons, and will work with quite row resolution / low amount of pixels per person.
Related
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 2 years ago.
Improve this question
For a research project at the institute I am working at, we are systematically collecting Street View Panoramas in certain areas.
In our country (Germany), a lot of buildings are censored. As I understand it, this is because according to our laws, Google must remove any personally identifying information upon request.
That is fine and I'm not looking to take away people's constitutional rights.
What I would like to be able to do is programmatically determine whether an image has one (or a certain percentage) of these blurred tiles in it, so we can exclude them as they are not useful to us.
I had a look at the metadata that I receive from a street view api request, but it did not look like there was such a parameter. Maybe I'm looking in the wrong place, though?
Thank you for your help :)
PS: "Alternative" solutions are also welcome - I have looked quickly into whether this kind of thing might be able to be done with certain image evaluation algorithms.
This might be a difficult/impossible task.
Blurred areas should have a lower noise amplitude, and you can enhance this by taking the gradient amplitude (possibly followed by equalization to increase contrast).
Anyway, real world images can also feature very uniform areas or slow shades, and if the image has low noise, there will be no way to distinguish them from blurred areas.
In addition, the images may be JPEG compressed, so that JPEG artefacts can be present and can strongly alter the uniformity and/or noise.
If a censored area is displayed as big pixels, then you have more luck: you can detect small squares of a uniform color, arranged in a grid. This never occurs in natural images. (But unfortunately again, lossy compression will make it harder.)
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 8 years ago.
Improve this question
I am doing a project which is hole detection in road. I am using a laser to emit beam on the road and using a camera to take a image of the road. the image may be like this
Now i want to process this image and give a result that is it straight or not. if it curve then how big the curve is.
I dont understand how to do this. i have search a lot but cant find a appropriate result .Can any one help me for that?
This is rather complicated and your question is very broad, but lets have a try:
Perhaps you have to identify the dots in the pixel image. There are several options to do this, but I'd smoothen the image by a blur filter and then find the most red pixels (which are believed to be the centers of the dots). Store these coordinates in a vector array (array of x times y).
I'd use a spline interpolation between the dots. This way one can simply get the local derivation of a curve touching each point.
If the maximum of the first derivation is small, the dots are in a line. If you believe, the dots belong to a single curve, the second derivation is your curvature.
For 1. you may also rely on some libraries specialized in image processing (this is the image processing part of your challenge). One such a library is opencv.
For 2. I'd use some math toolkit, either octave or a math library for a native language.
There are several different ways of measuring the straightness of a line. Since your question is rather vague, it's impossible to say what will work best for you.
But here's my suggestion:
Use linear regression to calculate the best-fit straight line through your points, then calculate the mean-squared distance of each point from this line (straighter lines will give smaller results).
You may need to read this paper, it is so interesting one to solve your problem
As #urzeit suggested, you should first find the points as accurately as possible. There's really no way to give good advice on that without seeing real pictures, except maybe: try to make the task as easy as possible for yourself. For example, if you can set the camera to a very short shutter time (microseconds, if possible) and concentrate the laser energy in the same time, the "background" will contribute less energy to the image brightness, and the laser spots will simply be bright spots on a dark background.
Measuring the linearity should be straightforward, though: "Linearity" is just a different word for "linear correlation". So you can simply calculate the correlation between X and Y values. As the pictures on linked wikipedia page show, correlation=1 means all points are on a line.
If you want the actual line, you can simply use Total Least Squares.
Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
We don’t allow questions seeking recommendations for books, tools, software libraries, and more. You can edit the question so it can be answered with facts and citations.
Closed 7 years ago.
Improve this question
Given N x-ray images with different exposure doses, I must combine them into a single one which condenses the information from the N source images. If my research is right, this problem falls in the HDRI cathegory.
My first approach is a weighted average. For starters, I'll work with just two frames.
Let A be the first image, which is the one with lowest exposure and thus is set to weigh more in order to highlight details. Let B be the second, overexposed image, C the resulting image and M the maximum possible pixel value. Thus, for each pixel i:
w[i] = A[i]/M
C = w[i] * A[i] + ( 1 - w[i] ) B[i]
An example result of applying this idea:
Notice how the result (third image) nicely captures the information from both source images.
The problem is that the second image has discontinuities around the object edges (this is unavoidable in overexposed images), and that carries on to the result. Looking closer...
The best reputed HDR software seems to be Photomatix, so I fooled around with it and no matter how I tweaked it, the discontinuities always appear in the result.
I think that I should somehow ignore the edges of the second image, but I must be do it in a "smooth way". I tried using a simple threshold but the result looks even worse.
What do you suggest? (only open source libraries welcome)
The problem here is that each image has a different exposure dose associated. Any HDR algorithm must take this into account.
I asked the people who created the x-ray images, and the exposure dose for the second image is approximately 4.2 times that of the first one. I was giving wrong EV values to Photomatix because I didn't know that EV is expressed in terms of stops, 1 stop meaning twice the reference value. So, assigning 0 EV to the first image and +2.1 EV to the second one, the discontinuities were gone, keeping all information.
Next problem was that I had no idea how Photomatix did this. So then I tried doing the same using Luminance HDR, aka qtpfsgui, which is open source.
To sum it up, the exposure bracketed images must be fed to an HDR compression algorithm, which creates an HDR image. Basically, that's a float point image which contains the information of all images. There are many algorithms to do this. Luminance HDR calls this HDR creation model and offers two of them: Debevec, and Robertson.
However, an HDR image cannot be displayed directly on a conventional display (i.e. monitor). So we need to convert it to a "normal" (LDR) image while keeping as much color information as possible. This is called tone-mapping, and there also various algorithms available for this; Luminance calls these Tonemap Operators and offers several. It also selects the most suitable one. The Pattanaik operator worked great for these images.
So now I'm reading Luminance's code in order to understand it and make my own implementation.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 6 years ago.
Improve this question
I am interested in recognizing letters on a Boggle board, probably using openCV. The letters are all the same font but could be rotated, so using a standard text recognition library is a bit of a problem. Additionally the M and W have underscores to differentiate them, and the Q is actually a Qu.
I am fairly confident I can isolate the seperate letters in the image, I am just wondering how to do the recognition part.
It depends on how fast you need to be.
If you can isolate the square of the letter and rotate it so that the sides of the square containing the letter are horizontal and vertical then I would suggest you:
convert the images to black/white (with the letter the one colour and the rest of the die the other
make a dataset of reference images of all letters in all four possible orientations (i.e. upright and rotated 90, 180 and 270 degrees)
use a template matching function such as cvMatchTemplate to find the best matching image from your dataset for each new image.
This will take a bit of time, so optimisations are possible, but I think it will get you a reasonable result.
If getting them in a proper orientation is difficult you could also generate rotated versions of your new input on the fly and match those to your reference dataset.
If the letters have different scale then I can think of two options:
If orientation is not an issue (i.e. your boggle block detection can also put the block in the proper orientation) then you can use the boundingbox of the area that has the letter colour as rough indicator of the scale of the incoming picture, and scale that to be the same size as the boundingbox on your reference images (this might be different for each reference image)
If orientation is an issue then just add scaling as a parameter of your search space. So you search all rotations (0-360 degrees) and all reasonable sizes (you should probably be able to guess a reasonable range from the images you have).
You can use a simple OCR like Tesseract. It is simple to use and is quite fast. You'll have to do the 4 rotations though (as mentioned in #jilles de wit's answer).
I made an iOS-app that does just this, based on OpenCV. It's called SnapSolve. I wrote a blog about how the detection works.
Basically, I overlay all 26x4 possible letters + rotations on each shape, and see which letter overlaps most. A little tweak to this is to smooth the overlay image, to get rid of artefacts where letters almost overlap but not quite.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 3 years ago.
Improve this question
I'm trying to develop a system, which recognizes various objects present in an image based on their primitive features like texture, shape & color.
The first stage of this process is to extract out individual objects from an image and later on doing image processing on each one by one.
However, segmentation algorithm I've studied so far are not even near perfect or so called Ideal Image segmentation algorithm.
Segmentation accuracy will decide how much better the system responds to given query.
Segmentation should be fast as well as accurate.
Can any one suggest me any segmentation algorithm developed or implemented so far, which won't be too complicated to implement but will be fair enough to complete my project..
Any Help is appreicated..
A very late answer, but might help someone searching for this in google, since this question popped up as the first result for "best segmentation algorithm".
Fully convolutional networks seem to do exactly the task you're asking for. Check the paper in arXiv, and an implementation in MatConvNet.
The following image illustrates a segmentation example from these CNNs (the paper I linked actually proposes 3 different architectures, FCN-8s being the best).
Unfortunately, the best algorithm type for facial recognition uses wavelet reconstruction. This is not easy, and almost all current algorithms in use are proprietary.
This is a late response, so maybe it's not useful to you but one suggestion would be to use the watershed algorithm.
beforehand, you can use a generic drawing(black and white) of a face, generate a FFT of the drawing---call it *FFT_Face*.
Now segment your image of a persons face using the watershed algorithm. Call the segmented image *Water_face*.
now find the center of mass for each contour/segment.
generate an FFT of *Water_Face*, and correlate it with the *FFT_Face image*. The brightest pixel in resulting image should be the center of the face. Now you can compute the distances between this point and the centers of segments generated earlier. The first few distances should be enough to distinguish one person from another.
I'm sure there are several improvements to the process, but the general idea should get you there.
Doing a Google search turned up this paper: http://www.cse.iitb.ac.in/~sharat/papers/prim.pdf
It seems that getting it any better is a hard problem, so I think you might have to settle for what's there.
you can try the watershed segmentation algorithm
also you can calculate the accuracy of the segmentation algorithm by the qualitative measures