Collinearity Equation

Objective: – Mathematically represent the general

relationship between corresponding image and ground
coordinates.

Concept: – Image Point, Object Point, and the

Photogrammetry Lect. #6 Perspective Center are collinear.
Collinearity equation and Exterior orientation
Asst. Prof. Dr.Haval AbdulJabbar Sadeq
Geomatics Eng. Dep., College of Engineering,
Salahaddin University-Erbil, Iraq.
2023/2024- Fall Semester
1 2

Collinearity Equation Derivation of the Collinearity Equation

 The functional model of the imaging system will be realized in the condition These image and object coordinate system will be related by three positional parameters and three orientation parameters, which can be
equation, 𝑥 𝑋−𝑋
 Condition equations are related to image points, object points, and imaging expressed as follow: 𝑦 = 𝑘𝑀 𝑌 − 𝑌
𝑧 𝑍−𝑍
system parameters.
 The condition equation is used for different purposes, depending on which 𝑥 − 𝑥𝑜 𝑋−𝑋
The value of the z is constant in the negative principle of distance: 𝑦 − 𝑦𝑜 = 𝑘𝑀 𝑌 − 𝑌
variables are considered observables, known, or unknowns in the stochastic −𝑓 𝑍−𝑍
model.
 An example is the collinearity condition equation, which may be used for 𝑥 − 𝑥𝑜 𝑚 𝑚 𝑚 𝑋−𝑋
By expressing the Matrix M in its elements then, we can get: 𝑦 − 𝑦𝑜 = 𝑘 𝑚 𝑚 𝑚 𝑌 − 𝑌 By multiplying the matrix by the
space resection, space intersection, relative orientation, and other tasks. −𝑓 𝑚 𝑚 𝑚 𝑍−𝑍
 The below figure is the geometry of a single point in a frame camera. As vector and eliminating the k factor, it is possible to obtain a classical form of the collinearity equations:
shown, the fundamental characteristic of frame imaging is that the 𝑚 𝑋 −𝑋 +𝑚 𝑌 −𝑌 +𝑚 𝑍 −𝑍
perspective centre, the image point and the corresponding object point all lie 𝑥 − 𝑥 = −𝑓.
𝑚 𝑋 −𝑋 +𝑚 𝑌 −𝑌 +𝑚 𝑍 −𝑍
on a line in space.
𝑚 𝑋 −𝑋 +𝑚 𝑌 −𝑌 +𝑚 𝑍 −𝑍
𝑦 − 𝑦 = −𝑓.
𝑚 𝑋 −𝑋 +𝑚 𝑌 −𝑌 +𝑚 𝑍 −𝑍
3 4
 Derivation of the Collinearity Equation Derivation of the Rotation Matrix
The above collinearity equations are used to determine the corresponding image coordinates when the interior, • Using the three rotation angles, the relationship between the image
exterior orientation and object point are known. space coordinate system (x, y, and z) and ground space coordinate
Also can be written as system (X, Y, and Z; or x’, y’, and z’) can be determined.

𝑚 𝑥−𝑥 +𝑚 𝑦−𝑦 + 𝑚 (−𝑓) • A 3 × 3 matrix defining the relationship between the two systems is
𝑋 − 𝑋 = (𝑍 − 𝑍 )
𝑚 𝑥−𝑥 +𝑚 𝑦−𝑦 + 𝑚 (−𝑓) used. This is referred to as the orientation or rotation matrix, M. The
rotation matrix can be defined as follows:
𝑚 𝑥−𝑥 +𝑚 𝑦−𝑦 +𝑚 −𝑓
𝑌−𝑌 = 𝑍−𝑍
𝑚 𝑥−𝑥 +𝑚 𝑦−𝑦 +𝑚 −𝑓 • To derive the rotation matrix M, three rotations are performed
The above collinearity equations are used to determine two components of object space (e.g. X and Y) given the sequentially: a primary rotation ω around the x-axis, followed by a
interior, exterior orientations, and image coordinates in addition to the rest component of the object space. secondary rotation ϕ around the y-axis, and a tertiary rotation κ around
the z-axis.

5 6

Derivation of the Rotation Matrix Collinearity Equation

• Each of the three elementary rotations is represented in matrix form as follows: By combining the above equations a relationship can be defined between the coordinates of an object point (P) relative
to the (X, Y, Z) and (𝑋 ,𝑌 ,𝑍 ): 𝑷 = 𝑴𝝎 × 𝑴𝝓 × 𝑴𝜿 × 𝑷𝝎𝝓𝜿

𝒎𝟏𝟏 𝒎𝟏𝟐 𝒎𝟏𝟑

By replacing 𝑀 × 𝑀 × 𝑀 𝑏𝑦 𝑀, M is a 3 × 3matrix: 𝑴 = 𝒎𝟐𝟏 𝒎𝟐𝟐 𝒎𝟐𝟑
𝒎𝟑𝟏 𝒎𝟑𝟐 𝒎𝟑𝟑

Where each entry of the matrix can be computed by equations :

Where x1,x2, and x3 are the coordinates before rotation and x’1,x’2 and x’3 are the coordinates after rotation.
Similarly, a rotation of ∅ about the x’2 axis and k about the x’’3 axis are given by.
𝒎𝟏𝟏 = 𝒄𝒐𝒔(∅) 𝒄𝒐𝒔 (𝜿)
𝒎𝟏𝟐 = 𝒔𝒊𝒏(𝝎 )𝒔𝒊𝒏 (𝝓)𝒄𝒐𝒔(𝜿) + 𝒄𝒐𝒔(𝝎)𝒔𝒊𝒏(𝜿)
𝒎𝟏𝟑 = −𝒄𝒐𝒔 𝝎 𝒔𝒊𝒏 𝝓 𝒄𝒐𝒔 𝜿 + 𝒔𝒊𝒏 𝝎 𝒔𝒊𝒏(𝜿)
𝒎𝟐𝟏 = −𝒄𝒐𝒔(∅) 𝒔𝒊𝒏 (𝜿)
𝒎𝟐𝟐 = −𝒔𝒊𝒏(𝝎)𝒔𝒊𝒏(∅) 𝒔𝒊𝒏 (𝜿) + 𝒄𝒐𝒔(𝝎)𝒄𝒐𝒔(𝜿)
𝒎𝟐𝟑 = 𝒄𝒐𝒔 𝒘 𝒔𝒊𝒏 𝝓 𝒔𝒊𝒏(𝜿) + 𝒔𝒊𝒏(𝝎)𝒄𝒐𝒔(𝜿)
𝒎𝟑𝟏 = 𝒔𝒊𝒏(∅)
𝒎𝟑𝟐 = −𝒔𝒊𝒏(𝝎) 𝒄𝒐𝒔(𝝓)
𝒎𝟑𝟑 = 𝒄𝒐𝒔 𝝎 𝒄𝒐𝒔(𝝓)
7 8
 Collinearity Equation-Example Exterior orientation
Example 1 For a photograph with exterior orientation elements (𝜔, 𝜙, 𝜅) = 2,5,15 Degrees and (𝑋 , 𝑌 , 𝑍 ) = (5000𝑚, 10,000𝑚, 2000𝑚)  Exterior orientation defines the camera's position and angular orientation at the time that the image is captured.
and camera parameters (𝑥 , 𝑦 , 𝑓) =(0.015, -0.0220, 152.4)mm, compute via the collinearity equations the coordinates of the ground point
 The variables defining the position and orientation of an image are referred to as the elements of exterior orientation.
(X,Y,Z)=(5100, 9800, 100)in the fiducial-based image system.
 The positional elements of exterior orientation include 𝜔, 𝜑, and, 𝜅, and the position of the perspective centre (O)
𝑚 𝑚 𝑚 0.9622 0.2616 −0.0751 Xo,Yo,and Zo, with respect to the ground space coordinate system (X, Y, and Z).
Solution: Evaluating the rotation matrix, using equation s yields:𝑀 = 𝑚 𝑚 𝑚 = −0.2578 0.9645 0.0562
𝑚 𝑚 𝑚 0.0871 −0.0348 0.9956  Zo is commonly referred to as the height of the camera above sea level, which is commonly defined by a datum.

Using equations of collinearity to obtain the image coordinates with respect to the [[principle point yields (x-xo, y-yo) =(15.159, -  Three rotation angles are shown below.
26.449)mm.

0.9622 5100 − 5000 + 0.2616 9800 − 10000 + −0.0751 100 − 2000

𝑥 − 𝑥 = −152.4. = 15.158𝑚𝑚
0.0871 5100 − 5000 + −0.0348 9800 − 10000 + 0.9956 100 − 2000
xa=15.158+0.015=15.173mm
. . .
. . .
:

ya =-26.440+(-0.0220)=-26.462mm
The coordinates in the fiducial-based system are (xa,ya)=(15.174, -26.469)mm.
9 10

Exterior orientation composites 1-Relative (Exterior) Orientation

Exterior orientation is of:  It is the process of creating a stereo model using a pair of images taken for the same
scene from different locations.
1-Relative (Exterior) Orientation
 Analytical relative orientation is the process of determining the relative angular
2-Absolute (exterior) orientation attitude and positional displacement between the photographs that existed when
photos were taken.

 It involves determining certain exterior orientation elements and calculating the

remaining ones.
 In analytical relative orientation, it is common practice to fix the exterior orientation
elements 𝜔, 𝜑, 𝜅, 𝑋𝐿 𝑎𝑛𝑑 𝑌𝐿 of the left photo of the stereo pair to zero values. The ZL
value of the left value is set equal to f, and XL of the right photo is equal to the photo
base b.
 The scale of the stereo model is approximately equal to the photo scale.

 The coplanarity condition equation can be used for analytical relative orientation, but
the collinearity condition is more commonly applied.

11 12
 Example 2-Absolute (exterior) orientation
Example: A stereopair of near-vertical photograph is taken with a It is the process of establishing the relationship between the model space and object space coordinates
152.114mm focal length camera. Photo coordinates of the images of six systems, so it can be ready for mapping and infer coordinates.
points in the overlap area are listed in the following table. Perform To achieve absolute orientations, seven parameters are involved,
analytical relative orientation of the stereo pair.  A uniform scale,
Solution: 1-With an ASCII text editor, create  Three translation, and,
2. Run the relor program to produce the
the following data file with a “.dat”  Three rotations.
following results:
extension: In the analytical stereo plotter, the absolute orientation is achieved by adding:
 Two horizontal control points in the model provide the scale, the translation along the X and Y
axes, and the k rotation around the Z-axis.
 Three elevation control points provide levelling information (the 𝜔rotation around x and 
rotation around y axes, and Z transition.
more points provide redundancy, which enables a least-square solution.
Once the transformation parameters have been computed, they can be applied to the remaining stereo
model points, including the XL, YL, and ZL coordinates of the left and right photographs. Thus operation
gives the coordinates of all stereo model points in the ground system.
13 14

2-Absolute (exterior) orientation: Example 2-Absolute (exterior) orientation

Example: Ground coordinates in a local vertical system for three control It is the process of establishing the relationship between the model space and object space coordinates
points are listed in the table below. For the result of the analytical relative systems, so it can be ready for mapping and infer coordinates.
orientation of the above example, perform analytical absolute orientation To achieve absolute orientations, seven parameters are involved,
using a three-dimensions; conformal coordinate transformation.
 A uniform scale,
Solution 1. With an ASCII text editor, create the 2. Run the “3dconf” program to produce the  Three translation, and,
following data file with a “.dat” extension: following results:  Three rotations.
In the analytical stereo plotter, the absolute orientation is achieved by adding:
 Two horizontal control points in the model provide the scale, the translation along the X and Y
axes, and the k rotation around the Z-axis.
 Three elevation control points provide levelling information (the 𝜔rotation around x and 
rotation around y axes, and Z transition.
Note: in the exterior orientation, the solution is more points provide redundancy, which enables a least-square solution.
found by using least-squares AX=L+V
Once the transformation parameters have been computed, they can be applied to the remaining stereo
A is the matrix of the coefficients, X is the
model points, including the XL, YL, and ZL coordinates of the left and right photographs. Thus operation
unknown and the L is a matrix of the constant
gives the coordinates of all stereo model points in the ground system.
term. V is the matrix of the residuals.
15 16