Planning for PhD? How about Remote Sensing?

Planning for PhD?

     If you are decided to do a Ph.D. in any computer science streams then this post is worth reading before starting your research.  In my experience, research gave me a new insight into looking at the problems and our surroundings. Only one thing that keeps me to move on all the impediments during my PhD is an enthralling research topic and specialization which I did chose. Before stepping into any research problems or literature the following three analysis you must do,

• What are your interests?
• In which stream you want to be specialized?
• Whether your interests/specialization is related to your guide stream? 

    Are you an earth explorer, nature enthusiast, agriculture endeavour or new on to the research? Then trust me 'Remote Sensing (RS)' domain could be your best choice to start up your research because it is one of the highlighted research areas with more interesting scopes especially with geographical measurements, Agriculture land classification, exploration and detection of the forest land cover areas and species, oceanography etc. Likewise, if you are interested in social networking then go for it without any second thoughts.

Remote Sensing

    Remote sensing domain interests me a lot because I am a nature lover, love to see more landscapes, and love to enjoy the best in seasons. Its always my dream to explore and travel all over the world. This fascination in me wants to understand our geographical structure and more onto the prediction. 

Image courtesy: Cervest (https://cervest.earth/remote-sensing-of-planet-earth-part-1-introduction-to-satellite-imagery/)

    As we all know, the studies on topography using remote sensing devices generate the enormous collection of rich spectral informative data with huge dimensionality for a field within a view. Interestingly, we rely mostly on the visual images to understand any pattern of the earth around us. Now, the advanced remote sensing instruments are capturing the images ahead of the human vision which has an excellent way to visualize the properties of the earth landscapes, vegetation, urban covering etc. 

    Basically, remote sensing instruments generate discrete or continuous wavelength bands.

    

 Remote Sensing Images: (a) Monochrome image contain a single band (b) RGB contains three discrete bands, (c) Spectroscopy with wavelength and reflectance, (d) Discrete multispectral bands (e) continuous hyperspectral bands

• Panchromatic images contain a single wavelength band,
• The discrete RGB contains the different wavelength bands of red, green & blue respectively
• Multispectral bands contain more than three bands which are discrete 
• Hyperspectral and Ultraspectral band contains hundreds of continuous narrow adjacent wavelength bands.

    The hyperspectral remote sensing images offer different spectral information for a particular scene.  Also, the hyperspectral data are complex in nature with more dimensionality and it contains rich spectral information for a particular scene. The hyperspectral data are collected through the imaging spectrometer. One interesting fact about human visible ranges in the electromagnetic spectrum is 380 nm to 700nm.  

 Electromagnetic Spectrum with different wavelengths in nanometers

Image Courtesy: https://www.colormatters.com/color-and-science/electromagnetic-color

 Finally! We are here at a conclusion

   First, choose your discipline based on your interests and scopes. As we all know, our earth is mostly surrounded by water, and most of the country's major occupations are agriculture, so it's worth trying on remote sensing domains. I chose hyperspectral data based on my interest, comfort zone, and also it is more challenging. This creates me to explore more on this for my PhD work. I strongly recommend the remote sensing domain, in particular, hyperspectral data which contains more research streams. I would like to share some interesting facts about the hyperspectral data classification paradigms and their applications in my next post.  Please give me more comments and say your thoughts about this article.

References:

1. Aggarwal, Charu C. Data mining: the textbook. Springer, 2015.
2. John B. Adams, and Alan R. Gillespie, “Remote Sensing of Landscapes with Spectral Images A Physical Modeling Approach”, 2006, Cambridge University Press, New York, ISBN-978-0-521-66221-5.
3. Marcus Borengasser, William S. Hungate, Russell Watkins, “Hyperspectral Remote sensing Principles and Applications”, 2008, Qihao Weng, Series Editor, CRC Press, Taylor and Francis group, ISBN-978-1-56670-654-4.
4. Shippert, Peg. "Introduction to hyperspectral image analysis." Online Journal of Space Communication 3.2003 (2003): 13.