Sunday, December 8, 2013

Image Processing: Improving Our View of the World

Image processing refers to the extraction on certain information from a 2D image by a computer. In Bioinformatics, a method called gel electrophoresis is used to perform analysis on DNA, RNA and proteins. An electrical charge in passed along gel that contain the DNA/RNA/protein(s) causing them to break apart and their fragments separate depending on their size and charge. By cutting the gel near where the tested molecules were, a black and white 2D image (2D page) shows the separation of the molecule parts. The 2D page is passed to a computer that analysis it by using an algorithm to compare it to another (known) 2D page. The analysis comes from corresponding spots with similar intensity (same shade). Through this type of testing, bioinformaticians are able to see how similar/difference the unknown molecule is to a known one.
Two 2D pages
Two 2D pages

Bioinformatics is just one field where image processing is used; with algorithms becoming more powerful, image processing is becoming more useful. Some algorithms are able to make images look clearer by removing “noise” from them (such as pictures from far away objects) while other are able to identify objects within the image – even if they are partly hidden.

3 comments:

  1. The application of image processing is so varied. I find it amazing that we can use computer software to analyze images, it saves so much time for scientists. The problem with image processing is that currently many things can trick computers into giving a false positive. While image processing might work for some applications images where there are any signs of segmentation or confusing patterns the computer fails. For example CAPTCHAs to humans seem so easy to solve but computers can't separate the letters from the noise.

    ReplyDelete
  2. Nice post Jonatan. Image processing is the wide field. There are various application of Image processing. Once such application is the volume rendering. This technique is used in the medical field. Volume rendering converts 3d image of the human body part to 2d image so it would be easy for the medicine practioner to analyze the image. There are more researches going on how to avoid problem of oversampling and uneven load balancing in the volume rendering technique.

    ReplyDelete
  3. Hello Jonatan,
    I thought it was very ambitious to try to cover image processing because it is such a general field. It was interesting to learn about gel electrophoresis, but since the post is very focused on that subject, I think that your title should include something more specific in relation to that. It would have been nice to know how they make images clearer as well. It was a very interesting read though; I had no idea what gel electrophoresis was! I hope to see more interesting things from you in the future.

    ReplyDelete