OCR and character similarity - math

I am currently working on some kind of OCR (Optical Character Recognition) system. I have already written a script to extract each character from the text and clean (most of the) irregularities out of it. I also know the font. The images I have now for example are:
M (http://i.imgur.com/oRfSOsJ.png (font) and http://i.imgur.com/UDEJZyV.png (scanned))
K (http://i.imgur.com/PluXtDz.png (font) and http://i.imgur.com/TRuDXSx.png (scanned))
C (http://i.imgur.com/wggsX6M.png (font) and http://i.imgur.com/GF9vClh.png (scanned))
For all of these images I already have a sort of binary matrix (1 for black, 0 for white). I was now wondering if there was some kind of mathematical projection-like formula to see the similarity between these matrices. I do not want to rely on a library, because that was not the task given to me.
I know this question may seem a bit vague and there are similar questions, but I'm looking for the method, not for a package and so far I couldn't find any comments regarding the method. The reason this question being vague is that I really have no point to start. What I want to do is actually described here on wikipedia:
Matrix matching involves comparing an image to a stored glyph on a pixel-by-pixel basis; it is also known as "pattern matching" or "pattern recognition".[9] This relies on the input glyph being correctly isolated from the rest of the image, and on the stored glyph being in a similar font and at the same scale. This technique works best with typewritten text and does not work well when new fonts are encountered. This is the technique the early physical photocell-based OCR implemented, rather directly. (http://en.wikipedia.org/wiki/Optical_character_recognition#Character_recognition)
If anyone could help me out on this one, I would appreciate it very much.

for recognition or classification most OCR's use neural networks
These must be properly configured to desired task like number of layers internal interconnection architecture , and so on. Also problem with neural networks is that they must be properly trained which is pretty hard to do properly because you will need to know for that things like proper training dataset size (so it contains enough information and do not over-train it). If you do not have experience with neural networks do not go this way if you need to implement it yourself !!!
There are also other ways to compare patterns
vector approach
polygonize image (edges or border)
compare polygons similarity (surface area, perimeter, shape ,....)
pixel approach
You can compare images based on:
histogram
DFT/DCT spectral analysis
size
number of occupied pixels per each line
start position of occupied pixel in each line (from left)
end position of occupied pixel in each line (from right)
these 3 parameters can be done also for rows
points of interest list (points where is some change like intensity bump,edge,...)
You create feature list for each tested character and compare it to your font and then the closest match is your character. Also these feature list can be scaled to some fixed size (like 64x64) so the recognition became invariant on scaling.
Here is sample of features I use for OCR
In this case (the feature size is scaled to fit in NxN) so each character has 6 arrays by N numbers like:
int row_pixels[N]; // 1nd image
int lin_pixels[N]; // 2st image
int row_y0[N]; // 3th image green
int row_y1[N]; // 3th image red
int lin_x0[N]; // 4th image green
int lin_x1[N]; // 4th image red
Now: pre-compute all features for each character in your font and for each readed character. Find the most close match from font
min distance between all feature vectors/arrays
not exceeding some threshold difference
This is partially invariant on rotation and skew up to a point. I do OCR for filled characters so for outlined font it may have use some tweaking
[Notes]
For comparison you can use distance or correlation coefficient

Related

Compress grayscale image using its histogram

I have a background on mathematics and Machine Learning, but I'm quite new on image compression. The other way I was thinking in the optimal way to compress an image just using a lookup table. This means, given an original image which has N unique values, change it to a new image with M unique values being M<N. Given a fixed value of M, my question was how to pick those values. I realized that if we take as figure of merit the total error (MSE) of all the pixels, all the information has to be in the histogram of the pixel intensities. Somehow, the most common values should be mapped to a closer value than the uncommon values, making the higher regions of the histogram more "dense" in the new values that the low regions.Hence I was wondering if it exists a mathematical formula that:
-Given the histogram h(x) of all the pixels intensities
-Given the number of uniques new values M
Defines the set of new M values {X_new} that minimizes the total error.
I tried to define the loss function and take the derivative, but it appeared some argmax operations that I don't know how to derivate them. However, my intution tells me that it should exist a closed formula.....
Example:
Say we have an image with just 10 pixels, with values {1,1,1,1,2,2,2,2,3,3}. We initially have N=3
and we are asked to select the M=2 unique values that minimizes the error. It is clear, that we have to pick the 2 most common ones, so {X_new}={1,2} and the new image will be "compressed" as {1,1,1,1,2,2,2,2,2,2}. If we are asked to pick M=1, we will pick {X_new}=2 to minimize the error.
Thanks!
This is called color quantization or palettization. It is essentially a clustering problem, usually in the 3D RGB space. Each cluster becomes a single color in the downsampled image. The GIF and PNG image formats both support palettes.
There are many clustering algorithms out there, with a lot of research behind them. For this, I would first try k-means and DBSCAN.
Note that palettization would only be one part of an effective image compression approach. You would also want to take advantage of both the spatial correlation of pixels (often done with a 2-D spatial frequency analysis such as a discrete cosine transform or wavelet transform), as well as taking advantage of the lower resolution of the human eye in color discrimination as opposed to grayscale acuity.
Unless you want to embark on a few years of research to improve the state of the art, I recommend that you use existing image compression algorithms and formats.

Best data structure & packages to represent geometric units on a grid

I want to write a program with 'geometry automata'. I'd like it to be a companion to a book on artistic designs. There will be different units, like the 'four petal unit' and 'six petal unit' shown below, and users and choose rulesets to draw unique patterns onto the units:
I don't know what the best data structure to use for this project is. I also don't know if similar things have been done and if so, using what packages or languages. I'm willing to learn anything.
All I know right now is 2D arrays to represent a grid of units. I'm also having trouble mathematically partitioning the 'subunits'. I can see myself just overlapping a bunch of unit circle formulas and shrinking the x/y domains (cartesian system). I can also see myself representing the curve from one unit to another (radians).
Any help would be appreciated.
Thanks!!
I can't guarantee that this is the most efficient solution, but it is a solution so should get you started.
It seems that a graph (vertices with edges) is a natural way to encode this grid. Each node has 4 or 6 neighbours (the number of neighbours matches the number of petals). Each node has 8 or 12 edges, two for each neighbour.
Each vertex has an (x,y) co-ordinate, for example the first row in in the left image, starting from the left is at location (1,0), the next node to its right is (3,0). The first node on the second row is (0,1). This can let you make sure they get plotted correctly, but otherwise the co-ordinate doesn't have much to do with it.
The trouble comes from having two different edges to each neighbour, each aligned with a different circle. You could identify them with the centres of their circles, or you could just call one "upper" and the other "lower".
This structure lets you follow edges easily, and can be stored sparsely if necessary in a hash set (keyed by co-ordinate), or linked list.
Data structure:
The vertices can naturally be stored as a 2-dimensional array (row, column), with the special characteristic that every second column has a horizontal offset.
Each vertex has a set of possible connections to those vertices to its right (upper-right, right, or lower right). The set of possible connections depends on the grid. Whether a connection should be displayed as a thin or a thick line can be represented as a single bit, so all possible connections for the vertex could be packed into a single byte (more compact than a boolean array). For your 4-petal variant, only 4 bits need storing; for the 6-petal variant you need to store 6 bits.
That means your data structure should be a 2-dimensional array of bytes.
Package:
Anything you like that allows drawing and mouse/touch interaction. Drawing the connections is pretty straightforward; you could either draw arcs with SVG or you could even use a set of PNG sprites for different connection bit-patterns (the sprites having partial transparency so as not to obscure other connections).

Handle "Division by Zero" in Image Processing (or PRNU estimation)

I have the following equation, which I try to implement. The upcoming question is not necessarily about this equation, but more generally, on how to deal with divisions by zero in image processing:
Here, I is an image, W is the difference between the image and its denoised version (so, W expresses the noise in the image), and K is an estimated fingerprint, gained from d images of the same camera. All calculations are done pixel-wise; so the equations does not involve a matrix multiplication. For more on the Idea of estimating digital fingerprints consult corresponding literature like the general wikipedia article or scientific papers.
However my problem arises when an Image has a pixel with value Zero, e.g. perfect black (let's say we only have one image, k=1, so the Zero gets not overwritten by the pixel value of the next image by chance, if the next pixelvalue is unequal Zero). Then I have a division by zero, which apparently is not defined.
How can I overcome this problem? One option I came up with was adding +1 to all pixels right before I even start the calculations. However this shifts the range of pixel values from [0|255] to [1|256], which then makes it impossible to work with data type uint8.
Other authors in papers I read on this topic, often do not consider values close the range borders. For example they only calculate the equation for pixelvalues [5|250]. They reason this, not because of the numerical problem but they say, if an image is totally saturated, or totally black, the fingerprint can not even be estimated properly in that area.
But again, my main concern is not about how this algorithm performs best, but rather in general: How to deal with divisions by 0 in image processing?
One solution is to use subtraction instead of division; however subtraction is not scale invariant it is translation invariant.
[e.g. the ratio will always be a normalized value between 0 and 1 ; and if it exceeds 1 you can reverse it; you can have the same normalization in subtraction but you need to find the max values attained by the variables]
Eventualy you will have to deal with division. Dividing a black image with itself is a proper subject - you can translate the values to some other range then transform back.
However 5/8 is not the same as 55/58. So you can take this only in a relativistic way. If you want to know the exact ratios you better stick with the original interval - and handle those as special cases. e.g if denom==0 do something with it; if num==0 and denom==0 0/0 that means we have an identity - it is exactly as if we had 1/1.
In PRNU and Fingerprint estimation, if you check the matlab implementation in Jessica Fridrich's webpage, they basically create a mask to get rid of saturated and low intensity pixels as you mentioned. Then they convert Image matrix to single(I) which makes the image 32 bit floating point. Add 1 to the image and divide.
To your general question, in image processing, I like to create mask and add one to only zero valued pixel values.
img=imread('my gray img');
a_mat=rand(size(img));
mask=uint8(img==0);
div= a_mat/(img+mask);
This will prevent division by zero error. (Not tested but it should work)

How to resize an existing point cloud file?

I am trying to enlarge a point cloud data set. Suppose I have a point cloud data set consisting of 100 points & I want to enlarge it to say 5 times. Actually I am studying some specific structure which is very small, so I want to zoom in & do some computations. I want something like imresize() in Matlab.
Is there any function to do this? What does resize() function do in PCL? Any idea about how can I do it?
Why would you need this? Points are just numbers, regardless whether they are 1 or 100, until all of them are on the same scale and in the same coordinate system. Their size on the screen is just a visual representation, you can zoom in and out as you wish.
You want them to be a thousandth of their original value (eg. millimeters -> meters change)? Divide them by 1000.
You want them spread out in a 5 times larger space in that particular coordinate system? Multiply their coordinates with 5. But even so, their visual representations will look exactly the same on the screen. The data remains basically the same, they will not be resized per se, they numeric representation will change a bit. It is the simplest affine transform, just a single multiplication.
You want to have finer or coarser resolution of your numeric representation? Or have different range? Change your data type accordingly.
That is, if you deal with a single set.
If you deal with different sets, say, recorded with different kinds of sensors and the numeric representations differ a bit (there are angles between the coordinate systems, mm vs cm scale, etc.) you just have to find the transformation from one coordinate system to the other one and apply it to the first one.
Since you want to increase the number of points while preserving shape/structure of the cloud, I think you want to do something like 'upsampling'.
Here is another SO question on this.
The PCL offers a class for bilateral upsampling.
And as always google gives you a lot of hints on this topic.
Beside (what Ziker mentioned) increasing allocated memory (that's not what you want, right?) or zooming in in visualization you could just rescale your point cloud.
This can be done by multiplying each points dimensions with a constant factor or using an affine transformation. So you can e.g switch from mm to m.
If i understand your question correctly
If you have defined your cloud like this
pcl::PointCloud<pcl::PointXYZ>::Ptr cloud (new pcl::PointCloud<pcl::PointXYZ>);
in fact you can do resize
cloud->points.resize (cloud->width * cloud->height);
Note that doing resize does nothing more than allocate more memory for variable thus after resizing original data remain in cloud. So if you want to have empty resized cloud dont forget to add cloud->clear();
If you just want zoom some pcd for visual puposes(i.e you cant see what is shape of cloud because its too small) why dont you use PCL Visualization and zoom by scrolling up/down

Similarity Between Colors

I'm writing a program that works with images and at some point I need to posterize the image. This means I need to bin the colors, but I'm having trouble deciding how to tell how close one color is to another.
Given a color in RGB, I can think of at least 2 ways to see how different they are:
|r1 - r2| + |g1 - g2| + |b1 - b2|
sqrt((r1 - r2)^2 + (g1 - g2)^2 + (b1 - b2)^2)
And if I move into HSV, I can think of other ways of doing it.
So I ask, ignoring speed, what is the best way to tell how similar two colors are? Best meaning most accurate to the human eye.
Well, if speed is not an issue, the most accurate way would be to take some sample images and apply the filter to them using various cutoff values for the distance (distance being determined by one of the equations on the Color_difference page that astander linked to, meaning you'd have to use one of those color spaces listed there with the calculations, then convert to sRGB or something [which also means that you'd need to convert the image into the other color space first if it's not in it to begin with]), and then have a large number of people examine the images to see what looks best to them, then go with the cutoff value for the images that the majority agrees looks best.
Basically, it's largely a matter of subjectiveness; in fact, it also depends on how stylized you want the images, and you might even want to add in some sort of control so that you can alter the cutoff distance on the fly.
If speed does become a bit of an issue and/or you want more simplicity, then just use your second choice for distance calculation (which is simply the CIE76 equation; just make sure to use the Lab* color space) with the cutoff being around 2 or 2.3.
What do you mean by "posterize the image"?
If you're trying to cluster the colors into bins, you should look at
cluster analysis
Just a comment if you are going to move to HSV (or similar spaces):
Diffing on H: difference between 0° and 359° is numerically big but perceptually is negligible.
H difference if V or S are small - is small.
For computer vision apps, more important not perceptual difference (used mostly by paint manufacturers) but are these colors belong to the same object/segment or not. Which means that we might partially ignore V, which can change from lighting conditions.

Resources