无标题文章

Edited by Jingchao Cao

GuId:

***

This is [an example](http://example.com/ "Title") inline link.

##Problem Set 1: Images as Functions

### Questions

#### 1. Input images

* ps1-2-a-1.png:

![ps1-1-a-1.png](file:///Users/cao/Graduate/Gatech/CV/Problems/ps1/input/ps1-1-a-1.png)

* ps1-1-a-2.png:

![ps1-2-a-2.png](file:///Users/

cao/Graduate/Gatech/CV/Problems/ps1/input/ps1-1-a-2.png)

#### 2. Color planes

* ps1-2-a-1.png:

![ps1-2-a-1.png](file:///Users/cao/Graduate/Gatech/CV/Problems/ps1/output/ps1-2-a-1.png)

* ps1-2-b-1.png:

![ps1-2-b-1.png](file:///Users/cao/Graduate/Gatech/CV/Problems/ps1/output/ps1-2-b-1.png)

* ps1-2-c-1.png:

![ps1-2-c-1.png](file:///Users/cao/Graduate/Gatech/CV/Problems/ps1/output/ps1-2-c-1.png)

* I think ps1-2-b-1.png (img1_green) looks more like a monochrome image.

#### 3. Replacement of pixels

* ps1-3-a-1.png:

![ps1-3-a-1.png](file:///Users/cao/Graduate/Gatech/CV/Problems/ps1/output/ps1-3-a-1.png)

#### 4. Arithmetic and Geometric operations

* 0 is the min of the pixel values of img1_green, and 255 is the          max. 132 is the mean. 69.52 is the standard deviation. First I use the function `flatten()` to change 2D array of img1_green into 1D array. Then I use functions `min()` and `max()`to calculate the min and max pixel values of img1_green. I compute mean pixel values by dividing the sum of every piexl value in flattened img1_green with the length of the array.

Print results:

__0 255 132 69.5151641486__

Code snippets:

img1_green_array = img1_green.flatten()

min_pixel = min(img1_green_array)

max_pixel = max(img1_green_array)

mean_pixel = sum(img1_green_array) / len(img1_green_array)

std = np.std(img1_green_array)

print min_pixel, max_pixel, mean_pixel, std

* ps1-4-b-1.png:

![ps1-4-b-1.png](file:///Users/cao/Graduate/Gatech/CV/Problems/ps1/output/ps1-4-b-1.png)

* ps1-4-c-1.png:

![ps1-4-c-1.png](file:///Users/cao/Graduate/Gatech/CV/Problems/ps1/output/ps1-4-c-1.png)

* ps1-4-d-1.png:

![ps1-4-d-1.png](file:///Users/cao/Graduate/Gatech/CV/Problems/ps1/output/ps1-4-d-1.png)

#### 5. Noise

* ps1-5-a-1.png:

![ps1-5-a-1.png](file:///Users/cao/Graduate/Gatech/CV/Problems/ps1/output/ps1-5-a-1.png)

The sigma is 30.

* ps1-5-b-1.png:

![ps1-5-b-1.png](file:///Users/cao/Graduate/Gatech/CV/Problems/ps1/output/ps1-5-b-1.png)

* The second one (ps1-5-b-1.png) looks better.

>

####这是一个普通的段落。

这是另一个普通段落。

AT

你可能感兴趣的:(无标题文章)