Update documentation image URLs

docs/Installation/Windows/Dragon_NaturallySpeaking.md

@@ -39,13 +39,13 @@ After installing Dragon, you can configure the DNS settings based on your prefer
 
 3. Configure the Natlink GUI
 
-   ![natlink_gui](/images/natlink_gui.png)
+   ![natlink_gui](https://github.com/dictation-toolbox/Caster/blob/master/docs/images/natlink_gui.png)
 
    - **Optionally** Check the relevant project Dragonfly or Unimacro to configure the file path to the grammars.
 
 4. Start Dragon start Dragon, the `Messages from Natlink` window should show loading a dragonfly script.  In the picture blow is an example loading module`_caster` is `_caster.py`.
 
-   ![natlink_running](/images/natlink_running.png)
+   ![natlink_running](https://github.com/dictation-toolbox/Caster/blob/master/docs/images/natlink_running.png)
 
 Scrips starting with an underscore and ending in .py `_*.py` will be imported in alphabetical order, except `__init__.py` will be loaded first if it exists. 
 

docs/Rule_Construction/Advanced_Caster_Rules/NodeRule.md

@@ -23,7 +23,7 @@ So, let's say for example that I want to be able to speak any of the following c
 
 That's sixteen specs getting added. Since a lot of them will only be spoken after others, this command set fits well into a tree structure. (Not all command sets will.) The number of specs in the active CCR grammar from this command set can be reduced by only keeping a certain number of levels of the tree structure open (speakable) at a given time.
 
-<img src="https://raw.githubusercontent.com/dictation-toolbox/Caster/master/docs/img/noderule1.png">
+<img src="https://github.com/dictation-toolbox/Caster/blob/master/docs/images/noderule1.png">
 
 In the above diagram, the first two levels of the tree are active. This means I can speak these specs:
 
@@ -41,7 +41,7 @@ In the above diagram, the first two levels of the tree are active. This means I
 
 If I then speak spec (a), it's effect occurs, and the NodeRule updates itself so that the active nodes look like the diagram below.
 
-<img src="https://raw.githubusercontent.com/dictation-toolbox/Caster/master/docs/img/noderule2.png">
+<img src="https://github.com/dictation-toolbox/Caster/blob/master/docs/images/noderule2.png">
 
 If I had spoken (a d) instead, only (m) would have been available next.
 

docs/Rule_Construction/Taxonomy_of_a_Rule.md

@@ -2,7 +2,7 @@
 
 Throughout the Caster documentation, you will see references to `rule`, `command`, `spec`, `action`, `extra`, and `default`s. These all refer to parts of a Rule, typically a MappingRule or MergeRule. Both are structured the same way, as depicted below.
 
-<img src="https://raw.githubusercontent.com/dictation-toolbox/Caster/master/docs/img/terminology.png">
+<img src="https://github.com/dictation-toolbox/Caster/blob/master/docs/images/terminology.png">
 
 - `spec` : This is what you say in order to invoke an `action`.
 - `action` : This is what happens when you speak a `spec`. Examples include Dragonfly's [Key](http://dragonfly.readthedocs.io/en/latest/actions.html#key-action), Text, and Function actions, or Caster's [R](http://caster.readthedocs.io/en/latest/caster/doc/readthedocs/ContextStack/#registeredaction), [ContextSeeker](http://caster.readthedocs.io/en/latest/caster/doc/readthedocs/ContextStack/#contextseeker), or [AsynchronousAction](http://caster.readthedocs.io/en/latest/caster/doc/readthedocs/ContextStack/#asynchronousaction) actions.