Look at this world through transform matrix | |
---|---|
To all viewers, This page is aimed to illustrate some phenomenons that we "see" in the perspectives of mathematics. Mainly, we would focus on astigmatism, pincushion and barrel distortion, as well as chromatic abberation. All these topics are very interesting to me, since they are always hanging around in our daily life. For example, it is easy to find astigmatism on some senior people. Maybe we would also suffer it later in our life. However, with the explanation of transformation matrix, we would learn more on how to tackle or correct it. Hope you enjoy this page!
- Joseph Ho
|
What's Transform Matrix?
Transform matrix consists of three types of matrix. They are (i)translation matrix, (ii)refraction matrix and (iii)reflection matrix. Multiplication of any combination among them would result a transform matrix which follows the progress of a ray through any optical system.
How can Astigmatism be corrected?
But, how's it related to mathematics? By transform matrix, we can understand and control the path of light. Cornea is a kind of lens made by human tissue. Astigmatism can then be illustrated from the point of view of mathematics, and further described by transform matrix. Moreover, It also provides the requierd data to produce correctness lens and even refractive surgery. Details... What's Barrel and Pincushion distortion?
What's Chromatic Aberration?
|
Transform Matrix |
---|
![]() Transforam matrix is also known as an ABCD matrix. Since it is a set of 2X2 matrix with A starts from the top left hand corner and B, C, D filled in it in a clockwise direction. A, B, C and D can be any real number, but the determinant of the matrix is always 1.
Translation matrix usually represents the ray pass through a media without changing angle. In other words, the ray is a straight line without bending within a media. Such ray tracing could be expressed by matrix with B equals the distance (S) the ray passed through in that media divided by the index of refraction (n) of that media. For instance, a ray passing through the air over 2 cm has B = 2 as refractive index of air is 1. In this case, all the values are unified by 'cm' in the matrix. (B = s/n)
Refraction matrix usually represents the ray pass through lens surface, in which the ray bends with respect to the index of refraction of the lens. Thus, C in the rafraction matrix will become -(n2-n1)/R. If the light ray comes from the left of the lens, n2 and n1 would be the index of refraction on the right media and the left media of the lens respectively. R is the radius of the lens, it has a negative value if the lens bend backward (with centre to the left of the surface). Reflection matrix represents the ray reflected usually by a mirror. This matrix is very similar to refraction matrix but involved only one media, resulting C = 2n/R. R is the radius of the mirror. |
Correction for Astigmatism |
---|
Eye ball
To understand how to correct astigmatism, we have to know more about the structure of eyes. In an eye ball, there's a cornea (C), a lens (L) with adjustable width controlled by a group of muscles atound it, and finally a place at the back called retina (R). It is a location where all light ray should reach.
If some of the rays by some reasons couldn't reach the retina, then image we see will become fuzzy.
Before we go straight to learn how to cure astigmatism by correction lens, we better know how to trace the rays by transform matrix. Since we learned that there are three types of matrix, they would then be our components or tools to build up the ray trace. For example, we would represent the rays in normal eye and astigmatism as following:
However, you may ask a question of whether we can simplify the calculation? This is a question a mathematician always ask for and the answer is "YES" in this case. Before we go to simplification, we have to make an assumption. It is the distance within the cornea and lens is very short. In other words, the cornea and lens are so thin that we can ignore the thickness. Thus, 's' is zero. This makes the translation matrix become an identity matrix. (i.e. A=D=1, B=C=0 ) Therefore, our new equation would be:
C = -1/f is the multiplication of two refraction matrixes, where f = (n2-n1)(1/R1-1/R2). R1 and R2 are the left and right refracting surface respectively. That means the R in M2 and M6 is R2, while the R in M4 and M8 is R1. Keep in mind that the above diagrams are only a bisection of an eye. But, at any angle we can view an eye from these bisection diagrams. With this key concept, we can further analyse astigmatism. Since the causes of astigmatism is due to the irregular shape of the cornea. This means there's a wrong value inside either M6, M7, M8 or all of them. This cue bring us a step further on how to tackle astignmatism and even some other eye-illnesses involved lens. Refraction surgery To allign with our assumption that cornea is a very thin substance of neglectable thickness, we may ignore M7 which is an identity matrix. This assumption leaves us two refraction matrixes, M6 and M8. We should first look at what variables involved in the matrixes and what they mean in our treatment. They are n1, n2 and R. 'n1' is the index of refraction of the fluid inside our eyes. 'n2' is the index of refraction of the media outside our eyes, which is air of index value 1. And the last one is R which is the curvature of our cornea. Obviously, we would not extract all the fuild out from our eyes and fill in another substance with an appropriate refraction index. And we cannot replace air with other things too. Thus, we may also neglect n1 and n2 which are the variables we cannot change. There are two R's left, one is the inside curvature of a cornea, and another one is the outside curvature. They are the only variables we may alter by refraction surgery. It alters the curvature of the cornea by using laser, which can only remove tissues from cornea. The extent of the tisses removed is determined by emmitting rays on the irregular part of cornea in order to see how far the rays apart from the retina. This is 'y', and surgeon would try to remove tissues to make it zero so that all rays could fall onto the retina again. (This is true, if we assume retina is horizontally parallel to the pupil, otherwise, we have to measure how far it's away from the pupil and set this value as the desired 'y'.)
This matrix is the multiplication result of the above matrix. The asterisks in the matrix are the values we may ignore in this case, since they do not affect the resulting 'y*'. To obtain 'y*' = 0, we have to make 's'='f'. Recall that f = (n2-n1)/(1/R1 - 1/R2). Thus, we may change the curvature that is R to alter 'f' in order to make it equals to 's'. Ha! This is our treatment to astignmatism by refraction surgery. Correction lens If you understand the idea behind refraction surgery, it's not hard to understand how correction lens resolve astigmatism. It is just place an appropriate lens in front of the eye-ball, like the following matrixes do.
= The last matrix is obtained by retrieving the * in the B's position of the pervious line matrix times the refraction matrix. It seems quite complicated, but the idea is simple and easy to solve. We always wants A=0, thus we have to solve nf^3 - f^3 - 2sf + s^2 = 0. This can be solved with only unknown f. Once we obtain f, we would then know the suitable curvature of the lens. This gives us the data to make an appropriate correction lens.
|
Barrel and pinchusion distortion |
---|
Chromatic Aberration |
---|