image-feature-enhancement.md 1.66 KB
Newer Older
Christian Tischer's avatar
Christian Tischer committed
1 2
# Image feature enhancement

Christian Tischer's avatar
Christian Tischer committed
3

Christian Tischer's avatar
Christian Tischer committed
4
<img src='https://g.gravizo.com/svg?
5
digraph G {
Christian Tischer's avatar
Christian Tischer committed
6 7
    shift [fontcolor=white,color=white];
    image -> filter -> "enhanced image";
8 9
    node [shape=box, color=grey, fontcolor=grey];
    "enhanced image" -> "feature" [label=" aka", style=dashed, color=grey, fontcolor=grey, fontsize=10];
10
    "feature enhancement" [shape=box, color=grey, fontcolor=grey, margin=0.05];
11
    filter -> "feature enhancement" [label=" aka", style=dashed, color=grey, fontcolor=grey, fontsize=10];   
Christian Tischer's avatar
Christian Tischer committed
12 13 14 15 16 17 18 19 20 21 22
}
'/>


## Difference of Gaussian (DoG) for spot enhancement

<img src='https://g.gravizo.com/svg?
 digraph G {
	shift [fontcolor=white,color=white];
	image -> "small blur";
	image -> "large blur";
Christian Tischer's avatar
Christian Tischer committed
23 24
	"small blur" -> "noise filtered";
	"large blur" -> "local background";
Christian Tischer's avatar
Christian Tischer committed
25 26 27
	"small blur" -> "small blur - large blur" -> "DoG";
	"large blur" -> "small blur - large blur" -> "DoG";
	"DoG" -> "Laplacian of Gaussian (LoG)" [label="  is related"];	
Christian Tischer's avatar
Christian Tischer committed
28 29 30 31 32 33
}
'/>

### Activity: Enhance spots in noisy image with uneven background

- Open image: xy_8bit__two_spots_noisy_uneven_background.tif
Christian Tischer's avatar
Christian Tischer committed
34 35 36 37 38 39 40
- Appreciate that you cannot readily threshold the spots
- Compute DoG:
	- Copy image and blur with a Gaussian of small sigma -> Gs
	- Copy image and blur with a Gaussian of bigger sigma -> Gb
		- For the official DoG: `rb = sqrt(2) * rs`
	- Create `DoG = Gs - Gb`
- Appreciate that now it is possible to threshold the spots in the DoG image
Christian Tischer's avatar
Christian Tischer committed
41
 
Christian Tischer's avatar
Christian Tischer committed
42 43
### Learn more

Christian Tischer's avatar
Christian Tischer committed
44
- https://imagescience.org/meijering/software/featurej/
Christian Tischer's avatar
Christian Tischer committed
45
- https://en.wikipedia.org/wiki/Difference_of_Gaussians
Christian Tischer's avatar
Christian Tischer committed
46
- https://github.com/CellProfiler/CellProfiler/blob/master/cellprofiler/modules/enhanceorsuppressfeatures.py#L4
Christian Tischer's avatar
Christian Tischer committed
47 48


Christian Tischer's avatar
Christian Tischer committed
49 50 51 52 53
### Formative Assessment

TODO