I did not took computer vision in college but I'd like to give it a try.
I want to make 3D model from a set of pictures.
I know you can do it with 123d catch or agisoft photoscan. doing it is not the point, it's writing the software.
At first I want to do stereo image matching then reconstruction from those 2 image, then multi view image matching and reconstruction.
According to this :
http://vision.middlebury.edu/stereo/eval/
The best algorythm is TSGO.
However I can't find any information about this TSGO algoritm, would any of you would know what TSGO stand for?
or if you know a better one.
Thanks.
The link on the page refers to a yet-to-be published paper:
[143] Anonymous. Accurate stereo matching by two step global
optimization. ECCV 2014 submission 74.
You'll have to wait until ECCV 2014 (September 6-12, 2014) to read it.
Meanwhile, you can take a look at OpenCV. It implements several stereo algorithms and get help you get up and running with the setup. Once you write your own implementation, you can contribute it to the community via OpenCV.
The paper and code is available at
http://www.cvc.uab.es/~mozerov/Stereo/index.htm
But it is not the best now...
Related
I'm trying to get the calibration values to put in the m210_stereo_param.yaml, like it's suggested in the official developer website (OnboardSDK for Linux). The objective is to have good values to test the depth perception sample. The website suggest different approach for the calibration, and I chose the OpenCV one.
I found an example of calibration on this Github repository: Opencv - stereo_calibrate_rc (some explanations are given on this link: Stereo Camera Calibration in Opencv 3)
The problem is that after getting the final Matrix (in the intrinsics.yaml and extrinsics.yaml), I modified the values in the m210_stereo_param.yaml, and tried to run the sample. I got this result (which is not correct, even the default values of the m210_stereo_param.yaml had a better result).
Do you have any idea of what is going wrong with the calibration ? It's quite complicated to found a clear approach to get values to put in the yaml.
Solve, the problem was that the xml file containing the list of images need to looks like this :
"data/left01.jpg"
"data/right01.jpg"
"data/left02.jpg"
"data/right02.jpg"
"data/left03.jpg"
...
And mine xml file was more like this :
"data/left01.jpg"
"data/left02.jpg"
"data/left03.jpg"
...
"data/right01.jpg"
"data/right02.jpg"
So, if you use this example, check if your xml file alternate left-right pictures.
I asked the the tech support of DJI and he told me that the M210 does not support the calibration. In this case, you should not have these problems. Just use the original one.
I was trying to get RGBDemo(mostly reconstructor) working with 2 logitech stereo cameras, but I did not figure out how to do it.
I noticed that there is a opencv grabber in nestk library and its header file is included in the reconstructor.cpp. Yet, when I try "rgbd-viewer --camera-id 0", it keeps looking for kinect.
My questions:
1. Is RGBDemo only working with kinect so far?
2. If RGBDemo can work with non-kinect stereo cameras, how do I do that?
3. If I need to write my own implementation for non-kinect stereo cameras, any suggestion on how to start?
Thanks in advance.
if you want to do it with non-kinect cameras. You don't even need stereo. There are algorithms now that are able to determine whether two images' viewpoints are sufficiently different that they can be used as if they were taken by a stereo camera. In fact, they use images from different cameras that are found on the internet and reconstruct 3D models of famous places. I can write you a tutorial on how to get it working. I've been meaning to do so. The software is called Bundler. Along with Bundler, people often also use CMVS and PMVS. CMVS preprocesses the images for PMVS. PMVS generates dense clouds.
BUT! I highly recommend that you don't go this route. It makes a lot of mistakes because there is so much less information in 2D images. It makes it very hard to reconstruct the 3D model. So, it ends up making a lot of mistakes, or not working. Although Bundler and PMVS are awesome compared to previous software, the stuff you can do with kinect is on a whole other level.
To use kinect will only cost you $80 for the kinect off of ebay or $99 off of amazon and another $5 for the power adapter off of amazon. So, I'd highly recommend this route. Kinect provides much more information for the algorithm to work with than 2D images do, making it much more effective, reliable and fast. In fact, it could take hours to process images with Bundler and PMVS. Whereas with kinect, I made a model of my desk in just a few seconds! It truly rocks!
I am working on a project to stitch together around 400 high resolution aerial images around 36000x2600 to create a map. I am currently using OpenCV and so far I have obtained the match points between the images. Now I am at a lost in figuring out how to get the matrix transformation of the images so I can begin the stitching process. I have absolutely no background in working with images nor graphics so this is a first time for me. Can I get some advice on how I would approach this?
The images that I received also came with a data sheet showing longitude, latitude, airplane wing angle, altitude, etc. of each image. I am unsure how accurate these data are, but I am wondering if I can use these information to perform the proper matrix transformation that I need.
Thanks
Do you want to understand the math behind the process or just have an superficial idea of whats going on and just use it?
The regular term for "image snitching" is image alignment. Feed google with it and you'll find tons of sources.
For example, here.
Best regards,
zhengtonic
In recent opencv 2.3 release...they implemented a whole process of image stitching. Maybe it is worth looking at.
Problem
Windows Mobile / Directdraw: Rotate video stream
The video preview is working, all I need now is a way to rotate the image. I think the only way to handle this is to write a custom filter based on CTransformFilter that will rotate the camera image for you. If you can help me to solve this problem, e.g. by helping me to develop this filter with my limited DirectDraw knowledge, the bounty is yours.
Background / Previous question
I'm currently developing an application for a mobile device (HTC HD2, Windows Mobile 6). One of things the program needs to do is to take pictures using the built-in camera. Previously I did this with the CameraCaptureDialog offered by the Windows Mobile 6 SDK, but our customer wants a more user-friendly solution.
The idea is to preview the camera's video stream in a control and take a high resolution picture (>= 2 megapixels) using the camera's photo function, when the control is clicked. We did some research on the topic and found out the best way to accomplish this seems to be using Direct Draw.
The downsides are that I never really used any native windows API and that my C++ is rather bad. In addition to this I read somewhere that the Direct Draw support of HTC phones is particularity bad and you will have to use undocumented native HTC libraries calls to take high quality pictures.
The good news is that a company offered us to develop a control that meets the specifications stated above. They estimated it would take them about 10 days, which lead to the discussion if we could develop this control ourself within a reasonable amount of time.
It's now my job to research which alternative is better. Needless to say it's far too less time to study the whole architecture and develop a demo, which lead me to the following questions:
Questions no longer relevant!
Does any of you have experience with similar projects? What are your recommendations?
Is there a good Direct Draw source code example that deals with video preview and image capturing?
Well if you look at the EZRGB24 sample you get the basics of a simple video transform filter.
There are 2 things you need to do to the sample to get it to do what you want.
1) You need to copy x,y to y,x.
2) You need to tell the media sample that the sample is now Height x Width instead of Width x Height.
Bear in mind that the final image will have exactly the same number of pixels.
To solve 1 is relatively simple. You can calculate the position of a pixel by doing "x + (y * Width)". So you step through each x and y calculate the position that way and then write it to "y + (x * Height)". This will transpose the image. Of course without step2 this will look completely wrong.
To solve 2 you need to get the AM_MEDIA_TYPE of the input sample. You then need to find out what the formatType is (Probably FormatType_VideoInfo or FormatType_VideoInfo2). You can thus cast the pbFormat member of AM_MEDIA_TYPE to either a VIDEOINFOHEADER or a VIDEOINFOHEADER2 (Depending on the FormatType). You need to now set VIDEOINFOHEADER[2]::bmiHeader.biWidth and biHeight to the biHeight and biWidth (respectively) of the input media sample. Everything else should be the same as the input AM_MEDIA_TYPE.
I hope that helps a bit.
This question will help you get some details about DirectDraw. I did some research about this some time ago and the best I could find was this blog post (also mentioned in the above question). The post presents an extension of the CameraCapture sample in the SDK.
However, don't have high expectations. It seems that the preview and the picture taken will only work in small resolution. Although DirectDraw does describe a way of configuring the resolution, there is no guarantee that this will be properly implemented by the driver.
So from my experience what you have read is true. The only way to do it will be to use HTC drivers. So, if you don't want to spend endless days in reverse engineering for a doubtful result, let someone else do the job for you. If you want to give it a shot, try xda-developers forum.
I am a newbie and i would like to give artistic effects to the images such as photocopy , blur ,glow edge, mosaic bubble ,pencil sketch ,marker , glass effect,paint brush,glow diffused effect programatically?
I want to implement it and any input is appreciated.
The following link depicts what i want to do exactly...
http://picasaweb.google.com/ashish.j.soni/BloggerPictures?authkey=Gv1sRgCMPHzP2P_fjRlgE#5413931800038472178
I believe that these effects are implemented in the GIMP.
The source code is available online http://www.gimp.org/source/ This may help you see how such algorithms are implemented.
You should check OpenCV, it will make your life easier. It is aimed at real-time computer vision, Not really what you are after, but it has a lot of useful functionality (eg. opening/saving different image formats) and standard image operations.
Matlab/octave is also a good tool to get feeling with image processing algorithms.