This should allow many algorithms to take advantage of more parallelization
technologies.

Also cv::, cv::gpu:: and cv::ocl:: namespace prefixes can be safely omitted
inside CV_ENUM and CV_FLAGS

initial optical flow was not cleaned during the second call

* added DenseOpticalFlow interface
* moved OpticalFlowDual_TVL1 to src folder

This fixes bug #2075

Invalid links and refences fixed.
SpaeseMat class documentation updated.

issue 2526

All errors were found by doc/check_docs2.py

added recommended number of stripes to parallel_for_, modified some of the functions to use larger stripes (for better performance)

+ Improve performance of calcOpticalFlowSingleScale method
+ Small refactoring

Current results:

IMAGE NAMES          RMSE
Beanbags
Dimetrodon           0.329428
DogDance
Grove2               0.550852
Grove3               1.464699
Hydrangea            0.523277
MiniCooper
RubberWhale          0.367246
Urban2               2.717003
Urban3               4.185070
Venus                0.775422
Walking

Time (for Urban3):
17.490248 sec

+ Fixed warnings
+ Add new function calcOpticalFlow with smaller number of arguments
+ Add asserts to algorithm and remove 'exit(1)'

+ New format for flow data - CV_32C2
+ Memory optimization
+ Cross Bilateral Filter optimization
+ Minor optimizations
+ Sample for calcOpticalFlowSF improved

"SimpleFlow" optical flow estimation algorithm (GSoC project) declaration in includes, implementation, usage example, test

+ Fixed warnings
+ Add new function calcOpticalFlow with smaller number of arguments
+ Add asserts to algorithm and remove 'exit(1)'

+ New format for flow data - CV_32C2
+ Memory optimization
+ Cross Bilateral Filter optimization
+ Minor optimizations
+ Sample for calcOpticalFlowSF improved

declaration in includes, implementation, usage example, test