# Image basics in mathematica

1) How the pixel values are defined in mathematica or how it is different from matlab?

2) I am wondering how image properties are different in mathematica, and how they are calculated? ComponentMeasurement: properties: such as "Centroid", "IntensityCentroid", "Area".

3) I want to find basic image processing link for mathematica 8 or 9?

I felt documentation is not informative enough, any link or help on this topic will be great.

• Mathematica does not "define" pixel values. Pixel values are defined by the image itself. Whoever creates an image, decides on the type of the image, and hence what format to store the image data in. This is no different than in Matlab. If you do ImageType[image] it will tell you the type of pixels of the image. byte, 16bit, Real, Bit, ..This is similar to Matlab's imfinfo() function. – Nasser Feb 25 '14 at 4:11
• @Nasser thanks for clarifying. – user12455 Feb 25 '14 at 4:15
• Another good reference is: reference.wolfram.com/mathematica/guide/ImageProcessing.html – bill s Feb 25 '14 at 5:20

1) How the pixel values are defined in mathematica, does each pixel value in image is weighted by it neighborhood configuration? or how it is different from matlab?

A pixel is just a number or a vector inside a 2d matrix. If it is a number, the image is in "Grayscale", if it is a vector like then there are several possibilities. The most used form is probably {r,g,b}, where the three entries define the red, the green, and the blue color component. Much information can be found in the Details and Options section here. An example where the vector is interpreted as hue color could be the following

Image@Table[{Sin[x], Cos[y]}, {x, 0, 2 Pi, .1}, {y, 0, 2 Pi, 0.1}]


I cannot say much about the difference to matlab, but in image processing in general the most used form of a pixel is just a number. Important is often how much bits are used to store the pixel number. The most common form is one byte per gray pixel (or 3 bytes for color).

2) I am wondering how image properties are different in mathematica, and how they are calculated? ComponentMeasurement: properties: such as "Centroid", "IntensityCentroid", "Area".

I'm struggling to understand what you want to know. Different to what? Properties of image components are by no means specific to Mathematica. They are known for a very long time and many of them are explained and implemented zillion times. About the specific algorithms used by Mathematica not much can be said. Something like Centroid is so easy to calculate that you shouldn't care about how exactly it is written down in Mathematica. For other, more complex image processing functions some details might be found in the "Details and Options" sections but in general those functions are black boxes.

3) I want to find basic image processing link for mathematica 8 or 9?

Now you lost me. There is no such thing like an image processing link. If you want to know more then the documentation is a very rich resource. Start here for an overview and work through the basic tutorial. This should give you a pretty good start.