Difference between revisions of "Team:Cambridge-JIC/MicroMaps"
K.armfield.1 (Talk | contribs) |
KaterinaMN (Talk | contribs) |
||
Line 92: | Line 92: | ||
<div style="width: 100%; padding:0% 10%; margin: 30px 0px;color:#fff"> | <div style="width: 100%; padding:0% 10%; margin: 30px 0px;color:#fff"> | ||
<h3>Image Processing</h3> | <h3>Image Processing</h3> | ||
+ | <!--*** ask ocean!!! *** | ||
+ | some things to consider: | ||
+ | cell counting | ||
+ | phenotype screening | ||
+ | fluorescence characterisation; relative measurement | ||
+ | identify samples (e.g. by shape and colour, maybe more complex features? e.g. eukaryotes -- look for nuclei) | ||
+ | ??? | ||
+ | |||
+ | *** ask souradip!!! *** | ||
+ | blockly visual programming to assemble simple annotators and microscope commands into complex workflows to automate experiments | ||
+ | --> | ||
+ | |||
+ | <p>The purpose of microscopy is to extract some useful information about the specimen: screen for a particular phenotype, examine fluorescence, measure sizes, count cells, recognize distinctive features, eg. nuclei. Focusing on a specimen is just a small part of the art of microscopy. The actual challenge is to interpret the image seen. Imagine a program that does this for you. This is what we had in mind when creating MicroMaps. To achieve this, we had to implement different types of image processing algorithms. | ||
+ | </p>Stages of image processing: | ||
+ | Macroscopic scale: - i.e. for use with shapeoko | ||
+ | Sample detection – the idea is to be able to detect points of interest (e.g. marchantia samples) with a wide angle camera that has a view of the entire stage. These points would then be stored and the the microscope used to take more detailed images of these. | ||
+ | There are 2 types of sample detection we implemented – standard thesholding, which makes an image greyscale and looks for the dark areas of the image. We started off with a basic contrast increase to isolate the darker areas of the image, which we assume would correspond to samples. Rajiv then followed the steps in a paper by ……… which was supposed to yield much better sample isolation for samples which look faint/are hard to distinguish from their background. This detected dents in agar gel. | ||
+ | The idea for colour detection implementation was to have an eye dropper to select the upper and lower colour darknesses to search for – the user would click to select these colours. These colours correspond to areas of the sample with better and worse illumination respectively. | ||
+ | The second type of sample detection implemented was colour detection. Into this was incorporated a slider that allows you to change the 'darkness' of the sample colour – which can vary depending on room lighting conditions. We tested this for marchantia gemma and it worked very well. Detected all samples before dents in agar jel. | ||
+ | Image stitching could be used with larger stages to create an image of the entire stage. | ||
+ | Microscopic scale: | ||
+ | Image stitching is used because we do not have variable zooming options so sometimes cannot see the entire sample in our field of view. We would be able to use this to create a full image of the sample. | ||
+ | The colour detection used above could be easily adapted to work with fluorescent samples – this would prove useful for sample counting and detection of, for example, samples that successfully show a specific fluorescent gene. | ||
+ | The colour detection could also be used in brightfield mode on samples with interestingly coloured features. | ||
+ | -STAINED SAMPLES? | ||
+ | </p> | ||
Revision as of 14:19, 18 September 2015