Activity 1: Digital Scanning
Still a little hung-up on a weekend where I got to experience a lot of my 'FIRSTs' (I'll let you dwell on what 'FIRSTs' was I able to experience... :P), I'm finally back to the real world, or am I?
From our weekend getaway, I did not bring any laptop with me to prevent myself from doing any schoolwork and enjoy the weekend fully, and now, I am back in front of my laptop to actually catch-up on a days worth of work.
NO LAPTOP for 2 DAYS! I can't even remember the last time I DID NOT use the computer. We are currently in the era where having a laptop (or internet connection) is almost a necessity since we make reports, graphs, and papers using a computer. It can also be seen that everything is computerized whether you want to enter a building, read and borrow books, buy stuff, or send important files to somewhere half across the world. As such, our writing materials (for publication, or plain writing) has evolved from a simple pen and paper to computer-based word processing.
Being a researcher dealing with large amount of data where there is a necessity to graph for interpretation, the technology already available enables easier production of such graphs. Imagine ourselves making graphs using a simple pen, and paper. Even the simple fitting of graphs would require extra materials such as the french curves which can be used to create drafts with smooth curves of almost any kind of curvature. [1] It was a challenge, even a chore, from our point of view so I say, aren't we very lucky???
To show the difference of plotting from the OLD TIMES (hahaha, sorry to those who are from this times) to computer-based plotting, we were asked to find a graph that was hand drawn, and scan it. The scanned image was then re-plotted (using computer) and this new plot was compared to the scanned image.
The scanned image was from the thesis of Minella G. Alarcon from the year 1986 with the title Optical remote sensing of gases using low-loss optical fiber line in the near-infrared region. It is a combination of two graphs: (A) Absorption of coefficient of CH4 molecules in city gas and (B) Received power/Noise power, both as function of spectral resolution (nm). This basically has two separate graphs superimposed.
The first thing to check was whether the image is tilted or not. So using Gimp 2, I adjusted the rotation of the image. After which, using Matlab, I determined the pixel values of ALL the necessary points.
Since obviously, the actual values of the points on the graph is not given, I need to find the relationship between the pixel value and physical value. So I related the physical value of the x-axis (tick points) (0,0.5,..) with that of the pixel value of x-axis using linear regression.This was done similarly with that of the y-axis. Therefore using these equations from the best fit lines (from linear regression), I simply substituted the pixel value to the equation (accordingly) and it gave the equivalent physical value.
Using the calculated physical values, I copied these to the MS Excel, plotted the points and superimposed the image. Superimposing the image with that of the graphs will show how accurate the reproduction of graphs was.
It is evident that there is a observable difference in the line fit using the French curves and using the computer.
To produce results faster and be able to produce the experiment in other images, I created a program which enabled me to do the same (until the calculation of actual values) for any image of choice. Application of the code to a different image is shown below. This is also from the same thesis.
One application of this technique is when we wish to determine the spectral response of a camera at a specific channel and wavelength. What is usually provided is a graph, not the actual values thus making it difficult to determine the equivalent values. Provided this technique, we can determine the needed values from the simple scan of the spectral graph.
One of the frustrations encountered when doing this activity is the actual tabulation of the pixel values of any point in the image, as well as the rotation of the image to ensure it is properly aligned in the x and y axis. I am very thankful that I was able to use several Matlab built-in functions to actually do the activity in a faster manner. Also, I am very thankful that I actually know how to use Gimp 2 (since I have no Adobe photoshop like most of my classmates have) so I was able to rotate the image.
I quite happy with the results.. For this activity, I give myself a 12. I understood the lesson and was able to produce all the required output (5). All the images and text are of good quality and the captions are stand alone (5). I also employed the technique with a different sample (2) as well as created a code that enables the reproduction of the activity on any desired image.
I would like to thank Phoebe Gallanosa for opening a blogspot account for me (hahaha, what a title!). Thank you Amihan! Thanks also to Art Galapon and Kirby Cheng's blogs for great insights and Francis Corpuz for letting me run on his laptop.
Reference/s:
[1] http://mathworld.wolfram.com/FrenchCurve.html
[2] Alarcon, Minella, Optical remote sensing of gases using low-loss optical fiber line in the near-infrared region, 1986
From our weekend getaway, I did not bring any laptop with me to prevent myself from doing any schoolwork and enjoy the weekend fully, and now, I am back in front of my laptop to actually catch-up on a days worth of work.
NO LAPTOP for 2 DAYS! I can't even remember the last time I DID NOT use the computer. We are currently in the era where having a laptop (or internet connection) is almost a necessity since we make reports, graphs, and papers using a computer. It can also be seen that everything is computerized whether you want to enter a building, read and borrow books, buy stuff, or send important files to somewhere half across the world. As such, our writing materials (for publication, or plain writing) has evolved from a simple pen and paper to computer-based word processing.
Being a researcher dealing with large amount of data where there is a necessity to graph for interpretation, the technology already available enables easier production of such graphs. Imagine ourselves making graphs using a simple pen, and paper. Even the simple fitting of graphs would require extra materials such as the french curves which can be used to create drafts with smooth curves of almost any kind of curvature. [1] It was a challenge, even a chore, from our point of view so I say, aren't we very lucky???
To show the difference of plotting from the OLD TIMES (hahaha, sorry to those who are from this times) to computer-based plotting, we were asked to find a graph that was hand drawn, and scan it. The scanned image was then re-plotted (using computer) and this new plot was compared to the scanned image.
 |
| Figure 1. Scanned image of the graphs [2] |
The first thing to check was whether the image is tilted or not. So using Gimp 2, I adjusted the rotation of the image. After which, using Matlab, I determined the pixel values of ALL the necessary points.
Since obviously, the actual values of the points on the graph is not given, I need to find the relationship between the pixel value and physical value. So I related the physical value of the x-axis (tick points) (0,0.5,..) with that of the pixel value of x-axis using linear regression.This was done similarly with that of the y-axis. Therefore using these equations from the best fit lines (from linear regression), I simply substituted the pixel value to the equation (accordingly) and it gave the equivalent physical value.
Using the calculated physical values, I copied these to the MS Excel, plotted the points and superimposed the image. Superimposing the image with that of the graphs will show how accurate the reproduction of graphs was.
 |
| Figure 2. Superimposed image and actual data plot for graph A |
 |
| Figure 3. Superimposed image and actual data plot for graph B |
It is evident that there is a observable difference in the line fit using the French curves and using the computer.
To produce results faster and be able to produce the experiment in other images, I created a program which enabled me to do the same (until the calculation of actual values) for any image of choice. Application of the code to a different image is shown below. This is also from the same thesis.
 |
| Figure 4. Scanned image of the graph (extra) |
 |
| Figure 5. Superimposed image and actual data plot for Figure 4 (extra) |
One application of this technique is when we wish to determine the spectral response of a camera at a specific channel and wavelength. What is usually provided is a graph, not the actual values thus making it difficult to determine the equivalent values. Provided this technique, we can determine the needed values from the simple scan of the spectral graph.
One of the frustrations encountered when doing this activity is the actual tabulation of the pixel values of any point in the image, as well as the rotation of the image to ensure it is properly aligned in the x and y axis. I am very thankful that I was able to use several Matlab built-in functions to actually do the activity in a faster manner. Also, I am very thankful that I actually know how to use Gimp 2 (since I have no Adobe photoshop like most of my classmates have) so I was able to rotate the image.
I quite happy with the results.. For this activity, I give myself a 12. I understood the lesson and was able to produce all the required output (5). All the images and text are of good quality and the captions are stand alone (5). I also employed the technique with a different sample (2) as well as created a code that enables the reproduction of the activity on any desired image.
I would like to thank Phoebe Gallanosa for opening a blogspot account for me (hahaha, what a title!). Thank you Amihan! Thanks also to Art Galapon and Kirby Cheng's blogs for great insights and Francis Corpuz for letting me run on his laptop.
 |
| Figure 6. Snapshot of the Matlab Code for the calculation of pixel-to-actual values |
Reference/s:
[1] http://mathworld.wolfram.com/FrenchCurve.html
[2] Alarcon, Minella, Optical remote sensing of gases using low-loss optical fiber line in the near-infrared region, 1986
Oh Mabel, double check your reference. I personnally know Dr. Minella Alarcn and she can't possibly have drawn those graphs in 1936! She's just in her 60's. It may have been 1986- probably the 8 got mistaken for a 3. Hahaha!
ReplyDeleteJust corrected it Ma'am. I was having difficulty discerning the title page because it was only captured via a cellphone camera. Gege just confirmed that it is indeed 1986.. Thanks Ma'am..
Delete