Run the hyper column on a given image with a given connection template. More...
#include <CINNIC/contourRun.H>
Public Member Functions | |
| void | copyCombinedSalMap (std::vector< Image< float > > &CSM) |
| copy the combined sal map to this class for analysis or dumping | |
| void | setArraySize (long size) |
| set the storage array size for linear components in subsq. iteration | |
| Image< float > | getSMI (int iter) |
| returned the processed salmap for this object at iter | |
| void | setIterations (int iter) |
| set number of iterations | |
| void | deriveEnergy () |
| find the time to total energy ratio | |
| void | dumpEnergySigmoid (const char *filename, const char *savefile, readConfig &config, Image< float > image, int scaleNo, int scaleTot) |
| Dump energy values computed with sigmoidal process. | |
| void | contourRunMain (std::vector< Image< float > > &imageMap, ContourNeuronCreate< float > &N, readConfig &config, Image< float > &group, int groups, int iter, float groupTop) |
| This is a short cut method to run runImage + other methods. | |
Public Attributes | |
| long | iterCounter |
Run the hyper column on a given image with a given connection template.
This is the central iterative process to CINNIC. It takes as input a 3D ImageMap and a 4D PropHold made from ContourNeuronCreate. It matches each neuron in the hyper column against each other neuron in the hyper column. Energy is tranfered to the other neuron based on a product of the two neurons current exitation from the ImageMap and the weight of thier connections. A negative weight signifies inhabition. Energys are allowed to build iteratively in an upper layer SalMap built into this class.
Definition at line 95 of file contourRun.H.
| void contourRun::contourRunMain | ( | std::vector< Image< float > > & | imageMap, | |
| ContourNeuronCreate< float > & | N, | |||
| readConfig & | config, | |||
| Image< float > & | group, | |||
| int | groups, | |||
| int | iter, | |||
| float | groupTop | |||
| ) |
This is a short cut method to run runImage + other methods.
This will run setConfig setImageSize, resetMatrix and runImage for you.
| Image | the processed image map of orientations from ImageMap | |
| N | This is the generic neuron template from ContourNeuronCreate | |
| config | This is the config file object from readConfig.C | |
| sizeX | The size of the image in X | |
| sizeY | The size of the image in Y |
Definition at line 535 of file contourRun.C.
| void contourRun::copyCombinedSalMap | ( | std::vector< Image< float > > & | CSM | ) |
copy the combined sal map to this class for analysis or dumping
Definition at line 77 of file contourRun.C.
| void contourRun::deriveEnergy | ( | ) |
find the time to total energy ratio
find the max energy per time slice, timestep is 1/x seconds where x = timestep
Definition at line 218 of file contourRun.C.
| void contourRun::dumpEnergySigmoid | ( | const char * | filename, | |
| const char * | savefile, | |||
| readConfig & | config, | |||
| Image< float > | image, | |||
| int | scaleNo, | |||
| int | scaleTot | |||
| ) |
Dump energy values computed with sigmoidal process.
| filename | The base name to give this file, should include path | |
| savefile | The tailend of the file name | |
| readConfig | the config file object |
Definition at line 145 of file contourRun.C.
References ASSERT, Image< T >::getHeight(), Image< T >::getWidth(), overlayStain(), rescale(), Image< T >::resize(), Image< T >::setVal(), sformat(), and Raster::WriteRGB().
| Image< float > contourRun::getSMI | ( | int | iter | ) |
returned the processed salmap for this object at iter
Definition at line 71 of file contourRun.C.
| void contourRun::setArraySize | ( | long | size | ) |
set the storage array size for linear components in subsq. iteration
Definition at line 115 of file contourRun.C.
| void contourRun::setIterations | ( | int | iter | ) |
1.6.3