In today’s digital era, computers have been essential tools not only in easy manipulation of data, but also in the restoration of old files and images. While it may seem so far-fetched, one can really restore pictures painted many centuries ago and produce a much cleaner version of the said image. Many professionals such as businessmen, doctors, engineers, and even artists benefit greatly from this.
Our first individual task in Applied Physics 186 is to do a digital scan of x-y graphs made decades ago. This is our first step in developing the skill of digital imaging. This task was very easy to do, although there was one hindrance in the work. I couldn’t find an old x-y graph! Old thesis files in my laboratory, which is the Plasma Physics Laboratory, do not have graphs that were made bare-handed or by even using old x-y plotters. Thesis books stored in the NIP library cannot be scanned and taken, so I felt very sad.
I tried opening our laboratory locker again, hoping for a miracle. In that blazing moment, I discovered an old CD inserted between the thesis files. The CD contains digital copies of the thesis files. Desperate for results, I opened the CD using my laptop. There I obtained a digital copy of an old thesis file. The paper was published in 1984 by K. Kadota, et. al., entitled Neutral Beam Probe Spectroscopy for Edge Plasma Diagnostics [1]. Apparently, our lab coordinator, Sir Henry Ramos, was a co-author in this paper. There, in that fleeting moment, I found the graph that would save me from utter despair. The graph contains data of the radial profiles of the electron excitation emission (Nυp) and the laser-induced fluorescence (Nυ1) for the Lithium resonance line. (That was a mouthful.) The picture is shown below.
I asked Ms. Eloi if this image is viable for the activity, and she said yes (Oh yeah!). So I was pumped up, and started the work with vigor and enthusiasm. The basic principle behind digital scanning is ratio and proportion. The first step was fairly easy. I used MS Paint to determine the pixel locations of the tick marks in the x and y axes. This is the first time I used image coordinates. In MS Paint, the first pixel coordinate is the column index, which increases from left to right. The second pixel coordinate is the row index, which increases from top to bottom.
Using the pixel coordinates, I can then measure the distance between two axis intervals via pixel counting. I used these values to create a pixel-to-actual value calibration graph, and ultimately, a calibration equation. The x and y calibration graphs, together with their equation, are posted below.
After obtaining these equations, I then chose a graph to digitally reproduce. I picked the (Nυ1) line (the one with hollow circles along with the line) for the activity. I was able to pick up 69 image points from the graph, which is also by using MS Paint. It was a bit tiring because you have to be very careful and precise in locating the coordinates. After choosing the points, the calibration equations were used to obtain the real values from the pixel values, and then VOILA!!!
I was able to reproduce the graph using only MS Paint and Excel! Moreover, our professor, Ma’am Jing, would give us bonus points if we are able to overlay the original graph with the digital image. I wanted the bonus points so I did it, and here’s the figure.
To do the overlay, one can save the original image as a separate image. One can then use the image to do a background overlay on the figure via MS Excel. Simply amazing! The digital image coincides almost perfectly with the original image, although there some slight deviation in the higher values. There were no problems with the resolution and alignment of the image in MS paint. The job was totally rewarding, and I am very happy with the results. It was amazing that one can actually restore these graphs with utmost perfection using computers. However, the task was a bit tedious since you have to manually locate the pixel coordinates. One could improve the digitizing process by creating a program that automatically detects the pixel coordinates via color, which is also not easy to do. Modern image scanners such as flatbed scanners found on work offices are based on the same technique discussed in this blog. However, these scanners utilize the amount of light intensity reflected by the paper to be scanned towards the CCD sensor (Charge-Coupled Device) [2]. The intensities received by the sensor are then converted from analog values to digital values. And thus, the scanned image can now be seen on the computer.
In this activity, I have fully understood all the concepts and technique involved. The result is great since the digital image coincides almost perfectly with the original image. If I am to grade myself, I would give myself 10/10 for this work. Oh wait! I almost forgot. I was able to do the additional bonus point, which is to overlay the original image with the digital graph using MS Excel. You know what that means.
Acknowledgements: I would like to thank the Plasma Physics Laboratory for this life-changing graph and Ms. Eloi, who verified that this graph is valid.
References:
[1] K. Kadota, et. al., “Neutral Beam Probe Spectroscopy for Edge Plasma Diagnostics,” Journal of Nuclear Materials 128 (1984): 960-964.
[2] Wikipedia Article on “Image Scanner.” Retrieved from: https://en.wikipedia.org/wiki/Image_scanner
Barteezy's Applied Physics 186 Experience 