1
$\begingroup$

Please help me tackle the below question:

Explain THREE forms of affine transformations using relevant examples for each case.

The answers I want to see are scaling, rotation and reflection.

$\endgroup$
6
$\begingroup$

Note: I have answered before the edit from trichoplax and I thought you were searching for other transformations other than the one you mentioned. The informations below are still useful so I will keep the answer here, but it does not directly answer your question.


Affine transformations (surprise!) map affine spaces to affine spaces. An affine space is substantially a vector space where you can establish an origin and define points as tuples of their coordinates. This is far from a formal definition, if you are interested in one I can edit the answer.

Now an affine transformation $T$ transform points into points and must preserve affine combinations:

$T(\lambda_1 P_1 + \lambda_2 P_2 + ... + \lambda_n P_n) = \lambda_1 T(P_1) + \lambda_2 T(P_2) + ... + \lambda_n T(P_n) $

Given $\sum_{i}^{n} \lambda_i = 1 $

However these transformations can't be arbitrary as the following must be preserved:

  • Parallelism is preserved. This means that if you transform parallel lines they remain parallel after the affine transformation.

  • Relative ratios are preserved. This means that if you have $R = (1-\beta) P + \beta Q$ then $T(R) = (1 - \beta) T(P) + \beta T(Q)$

  • Addition between vector and points are preserved. Meaning that $T( P + \vec{v} ) = T(P) + T(\vec{v})$

With these properties in mind you can come up with a very big number of affine transforms yourself. A couple of obvious ones other the ones you mentioned are:

Translation

$$T_vp = \begin{bmatrix} 1&0&0&v_x\\0&1&0&v_y\\0&0&1&v_z\\0&0&0&1\end{bmatrix} \begin{bmatrix} p_x+v_x\\p_x+v_y\\p_z+v_z\\1\end{bmatrix}=p+v$$

That moves a point into a specific direction by a specific amount specified by a displacement vector.

Shearing

enter image description here image from wikipedia

That is a transform that displace all points in a given direction by an amount that is proportional to their perpendicular distance to a line parallel to that direction.

For example the transform matrix for an horizontal shear in a 2D space is given by:

$$ \begin{bmatrix}x'\\y'\end{bmatrix} = \begin{bmatrix}x+my\\y\end{bmatrix} = \begin{bmatrix}1&m\\0&1\end{bmatrix}\begin{bmatrix}x\\y\end{bmatrix}$$


Again, these two are just example, the important information you should really keep is the definition (and properties) of an affine transform; with that in mind it shouldn't be too hard to recognize an affine transform. Also note that combining affine transforms will give you an affine transform!

$\endgroup$
  • 1
    $\begingroup$ Note that I might have answered a question that is slightly different from yours, but I did that on purpose as yours is clearly an homework question and I preferred to give you the tools to come up with an answer by yourself rather than giving you one directly :) $\endgroup$ – cifz Nov 13 '15 at 9:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.