Counting Objects via Mean-Shift Segmentation - algorithm

I'm trying to use mean-shift segmentation to count the objects found in an image. I've been working with [pyrMeanShiftFiltering][1] in OpenCV. Using the following code, I'm able to produce an image that has been segmented. However, I do not know how to actually count the number of "items" in that image.
Simply running pyrMeanShiftFiltering( img, res, spatialRad, colorRad, maxPyrLevel ); on this image
produces this image
In this example, it doesn't seem much different, although there are some images where the segmentation makes a huge difference in the colors and such present. However, for the majority of test cases, I'm going to assume that the colors will not be terribly distinct and that the edges will not be distinct (as they are in the example given) enough to use edge detection on the image itself.
Based off of this, how can I go about finding the number of objects found inside of that image? I'm looking for a bit of code, although any poke in the correct direction will help.

If the objects contain different colors (and those colors are distinctive) the simplest soultion would be to count how many clusters are there (remove cluster for the white color since the pages are white/yellow).
On images that you have shown you could also use corner detector since you have very distinctive corners, then look the surrounding and filter those corners by color (the surrounding color has to contain some color and white one (from pages)) then match corners that lie in the same vertical line and finally count them.
One other idea is to extrapolate white/yellow color from pages (clustering + histogram filtering) and to count different blobs.
Maybe the best approach is to extrapolate white/yeelow color from pages by finding cover color (blue, green) and closest white color => find blobs. Those blobs can be labeled to the closest cover color. Then you have blobs that present pages and are labeled according the closest cover color.
Those blobs may be broken to several pieces (one book partly covers the other book) and two different blobs may belong to the same object (a book with multiple color cover)
but you now that those blobs have to be rectangular. So you could find lines in that binarized image and try to connect them with closest line. Then you will finally have one blob that matches one book. Finally you can count them.

Related

Using Image Detection to detect lines?

I am trying to figure out how to use RANSAC for line detection, I am using canny edge detection to detect initial edges within the image, however would I randomly take a sample of the pixel values, or pixel locations from the image to model the lines.
You use locations of course. You cannot get a line from intensities.
Edge detectors usually generate two kinds of pixels. Non-edge pixels who are usually black and edge-pixels who are not black. Their values sometimes bear information on the edge quality.
Just pick a random set of edge-points...
Also there is plenty of literature online. Just google and read what you find.
http://cs.gmu.edu/~kosecka/cs682/lect-fitting.pdf for example...

anyway to remove algorithmically discolorations from aerial imagery

I don't know much about image processing so please bear with me if this is not possible to implement.
I have several sets of aerial images of the same area originating from different sources. The pictures have been taken during different seasons, under different lighting conditions etc. Unfortunately some images look patchy and suffer from discolorations or are partially obstructed by clouds or pix-elated, as par example picture1 and picture2
I would like to take as an input several images of the same area and (by some kind of averaging them) produce 1 picture of improved quality. I know some C/C++ so I could use some image processing library.
Can anybody propose any image processing algorithm to achieve it or knows any research done in this field?
I would try with a "color twist" transform, i.e. a 3x3 matrix applied to the RGB components. To implement it, you need to pick color samples in areas that are split by a border, on both sides. You should fing three significantly different reference colors (hence six samples). This will allow you to write the nine linear equations to determine the matrix coefficients.
Then you will correct the altered areas by means of this color twist. As the geometry of these areas is intertwined with the field patches, I don't see a better way than contouring the regions by hand.
In the case of the second picture, the limits of the regions are blurred so that you will need to blur the region mask as well and perform blending.
In any case, don't expect a perfect repair of those problems as the transform might be nonlinear, and completely erasing the edges will be difficult. I also think that colors are so washed out at places that restoring them might create ugly artifacts.
For the sake of illustration, a quick attempt with PhotoShop using manual HLS adjustment (less powerful than color twist).
The first thing I thought of was a kernel matrix of sorts.
Do a first pass of the photo and use an edge detection algorithm to determine the borders between the photos - this should be fairly trivial, however you will need to eliminate any overlap/fading (looks like there's a bit in picture 2), you'll see why in a minute.
Do a second pass right along each border you've detected, and assume that the pixel on either side of the border should be the same color. Determine the difference between the red, green and blue values and average them along the entire length of the line, then divide it by two. The image with the lower red, green or blue value gets this new value added. The one with the higher red, green or blue value gets this value subtracted.
On either side of this line, every pixel should now be the exact same. You can remove one of these rows if you'd like, but if the lines don't run the length of the image this could cause size issues, and the line will likely not be very noticeable.
This could be made far more complicated by generating a filter by passing along this line - I'll leave that to you.
The issue with this could be where there was development/ fall colors etc, this might mess with your algorithm, but there's only one way to find out!

Invoice / OCR: Detect two important points in invoice image

I am currently working on OCR software and my idea is to use templates to try to recognize data inside invoices.
However scanned invoices can have several 'flaws' with them:
Not all invoices, based on a single template, are correctly aligned under the scanner.
People can write on invoices
etc.
Example of invoice: (Have to google it, sadly cannot add a more concrete version as client data is confidential obviously)
I find my data in the invoices based on the x-values of the text.
However I need to know the scale of the invoice and the offset from left/right, before I can do any real calculations with all data that I have retrieved.
What have I tried so far?
1) Making the image monochrome and use the left and right bounds of the first appearance of a black pixel. This fails due to the fact that people can write on invoices.
2) Divide the invoice up in vertical sections, use the sections that have the highest amount of black pixels. Fails due to the fact that the distribution is not always uniform amongst similar templates.
I could really use your help on (1) how to identify important points in invoices and (2) on what I should focus as the important points.
I hope the question is clear enough as it is quite hard to explain.
Detecting rotation
I would suggest you start by detecting straight lines.
Look (perhaps randomly) for small areas with high contrast, i.e. mostly white but a fair amount of very black pixels as well. Then try to fit a line to these black pixels, e.g. using least squares method. Drop the outliers, and fit another line to the remaining points. Iterate this as required. Evaluate how good that fit is, i.e. how many of the pixels in the observed area are really close to the line, and how far that line extends beyond the observed area. Do this process for a number of regions, and you should get a weighted list of lines.
For each line, you can compute the direction of the line itself and the direction orthogonal to that. One of these numbers can be chosen from an interval [0°, 90°), the other will be 90° plus that value, so storing one is enough. Take all these directions, and find one angle which best matches all of them. You can do that using a sliding window of e.g. 5°: slide accross that (cyclic) region and find a value where the maximal number of lines are within the window, then compute the average or median of the angles within that window. All of this computation can be done taking the weights of the lines into account.
Once you have found the direction of lines, you can rotate your image so that the lines are perfectly aligned to the coordinate axes.
Detecting translation
Assuming the image wasn't scaled at any point, you can then try to use a FFT-based correlation of the image to match it to the template. Convert both images to gray, pad them with zeros till the originals take up at most 1/2 the edge length of the padded image, which preferrably should be a power of two. FFT both images in both directions, multiply them element-wise and iFFT back. The resulting image will encode how much the two images would agree for a given shift relative to one another. Simply find the maximum, and you know how to make them match.
Added text will cause no problems at all. This method will work best for large areas, like the company logo and gray background boxes. Thin lines will provide a poorer match, so in those cases you might have to blur the picture before doing the correlation, to broaden the features. You don't have to use the blurred image for further processing; once you know the offset you can return to the rotated but unblurred version.
Now you know both rotation and translation, and assumed no scaling or shearing, so you know exactly which portion of the template corresponds to which portion of the scan. Proceed.
If rotation is solved already, I'd just sum up all pixel color values horizontally and vertically to a single horizontal / vertical "line". This should provide clear spikes where you have horizontal and vertical lines in the form.
p.s. Generated a corresponding horizontal image with Gimp's scaling capabilities, attached below (it's a bit hard to see because it's only one pixel high and may get scaled down because it's > 700 px wide; the url is http://i.stack.imgur.com/Zy8zO.png ).

How do I locate black rectangles in a grid and extract the binary code from that

i'm working in a project to recognize a bit code from an image like this, where black rectangle represents 0 bit, and white (white space, not visible) 1 bit.
Somebody have any idea to process the image in order to extract this informations? My project is written in java, but any solution is accepted.
thanks all for support.
I'm not an expert in image processing, I try to apply Edge Detection using Canny Edge Detector Implementation, free java implementation find here. I used this complete image [http://img257.imageshack.us/img257/5323/colorimg.png], reduce it (scale factor = 0.4) to have fast processing and this is the result [http://img222.imageshack.us/img222/8255/colorimgout.png]. Now, how i can decode white rectangle with 0 bit value, and no rectangle with 1?
The image have 10 line X 16 columns. I don't use python, but i can try to convert it to Java.
Many thanks to support.
This is recognising good old OMR (optical mark recognition).
The solution varies depending on the quality and consistency of the data you get, so noise is important.
Using an image processing library will clearly help.
Simple case: No skew in the image and no stretch or shrinkage
Create a horizontal and vertical profile of the image. i.e. sum up values in all columns and all rows and store in arrays. for an image of MxN (width x height) you will have M cells in horizontal profile and N cells in vertical profile.
Use a thresholding to find out which cells are white (empty) and which are black. This assumes you will get at least a couple of entries in each row or column. So black cells will define a location of interest (where you will expect the marks).
Based on this, you can define in lozenges in the form and you get coordinates of lozenges (rectangles where you have marks) and then you just add up pixel values in each lozenge and based on the number, you can define if it has mark or not.
Case 2: Skew (slant in the image)
Use fourier (FFT) to find the slant value and then transform it.
Case 3: Stretch or shrink
Pretty much the same as 1 but noise is higher and reliability less.
Aliostad has made some good comments.
This is OMR and you will find it much easier to get good consistent results with a good image processing library. www.leptonica.com is a free open source 'C' library that would be a very good place to start. It could process the skew and thresholding tasks for you. Thresholding to B/W would be a good start.
Another option would be IEvolution - http://www.hi-components.com/nievolution.asp for .NET.
To be successful you will need some type of reference / registration marks to allow for skew and stretch especially if you are using document scanning or capturing from a camera image.
I am not familiar with Java, but in Python, you can use the imaging library to open the image. Then load the height and the widths, and segment the image into a grid accordingly, by Height/Rows and Width/Cols. Then, just look for black pixels in those regions, or whatever color PIL registers that black to be. This obviously relies on the grid like nature of the data.
Edit:
Doing Edge Detection may also be Fruitful. First apply an edge detection method like something from wikipedia. I have used the one found at archive.alwaysmovefast.com/basic-edge-detection-in-python.html. Then convert any grayscale value less than 180 (if you want the boxes darker just increase this value) into black and otherwise make it completely white. Then create bounding boxes, lines where the pixels are all white. If data isn't terribly skewed, then this should work pretty well, otherwise you may need to do more work. See here for the results: http://imm.io/2BLd
Edit2:
Denis, how large is your dataset and how large are the images? If you have thousands of these images, then it is not feasible to manually remove the borders (the red background and yellow bars). I think this is important to know before proceeding. Also, I think the prewitt edge detection may prove more useful in this case, since there appears to be less noise:
The previous method of segmenting may be applied, if you do preprocess to bin in the following manner, in which case you need only count the number of black or white pixels and threshold after some training samples.

How does Google's image color search work?

Let's say I query for
http://images.google.com.sg/images?q=sky&imgcolor=black
and I get all the black color sky, how actually does the algorithm behind work?
Based on this paper published by Google engineers Henry Rowley, Shumeet Baluja, and Dr. Yushi Jing, it seems the most important implication of your question about recognizing colors in images relates to google's "saferank" algorithm for pictures that can detect flesh-tones without any text around it.
The paper begins by describing by describing the "classical" methods, which are typically based on normalizing color brightness and then using a "Gaussian Distribution," or using a three-dimensional histogram built up using the RGB values in pixels (each color is a 8bit integer value from 0-255 representing how much . of that color is included in the pixel). Methods have also been introduced that rely on properties such as "luminance" (often incorrectly called "luminosity"), which is the density of luminous intensity to the naked eye from a given image.
The google paper mentions that they will need to process roughly 10^9 images with their algorithm so it needs to be as efficient as possible. To achieve this, they perform the majority of their calculations on an ROI (region of interest) which is a rectangle centered in the image and inset by 1/6 of the image dimensions on all sides. Once they've determined the ROI, they have many different algorithms that are then applied to the image including Face-Detection algs, Color Constancy algs, and others, which as a whole find statistical trends in the image's coloring and most importantly find the color shades with the highest frequency in the statistical distribution.
They use other features such as Entropy , Edge-Detection, and texture-definitions to
In order to extract lines from the images, they use the OpenCV implementation (Bradski, 2000) of the probabilistic Hough transform (Kiryati et al., 1991) computed on the edges of the skin color connected components, which allows them to find straight lines which are probably not body parts and additionally allows them to better determine which colors are most important in an image, which is a key factor in their Image Color Search.
For more on the technicalities of this topic including the math equations and etc, read the google paper linked to in the beginning and look at the Research section of their web site.
Very interesting question and subject!
Images are just pixels. Pixels are just RGB values. We know what black is in RGB, so we can look for it in an image.
Well, one method is, in very basic terms:
Given a corpus of images, determine the high concentrations of a given color range (this is actually fairly trivial), store this data, index accordingly (index the images according to colors determined from the previous step). Now, you have essentially the same sort of thing as finding documents containing certain words.
This is a very, very basic description of one possible method.
There are various ways of extracting color from an image, and I think other answers addressed them (K-Means, distributions, etc).
Assuming you have extracted the colors, there are a few ways to search by color. One slow, but obvious approach would be to calculate the distance between the search color and the dominant colors of the image using some metric (e.g. Color Difference), and then weight the results based on "closeness."
Another, much faster, approach would be to essentially downscale the resolution of your color space. Rather than deal with all possible RGB color values, limit the extraction to a smaller range like Google does (just Blue, Green, Black, Yellow, etc). Then the user can search with a limited set of color swatches and calculating color distance becomes trivial.

Resources