cvac::DetectorProperties Struct Reference

Public Attributes
int	verbosity
bool	isSlidingWindow = false
Size	nativeWindowSize
Size	slideStartSize
Size	slideStopSize
float	slideScaleFactor = MINUSONE
float	slideStepX = MINUSONE
float	slideStepY = MINUSONE
bool	canSetSensitivity = false
double	falseAlarmRate = -1.0
double	recall = -1.0
bool	canPostProcessNeighbors
int	minNeighbors = MINUSONE
double	videoFPS = -1.0
Properties	props

Detailed Description

Structure to obtain and set properties of detector services

Member Data Documentation

bool cvac::DetectorProperties::canPostProcessNeighbors

bool cvac::DetectorProperties::canSetSensitivity = false

double cvac::DetectorProperties::falseAlarmRate = -1.0

To use current values for one parameter, use the value of -1

bool cvac::DetectorProperties::isSlidingWindow = false

int cvac::DetectorProperties::minNeighbors = MINUSONE

-1 for default

Size cvac::DetectorProperties::nativeWindowSize

Returns the size of the "native" detector, that is, the size of objects that can be detected without scaling

Properties cvac::DetectorProperties::props

Props are name-value pairs that can specify detector-specific properties, if any.

double cvac::DetectorProperties::recall = -1.0

float cvac::DetectorProperties::slideScaleFactor = MINUSONE

Size cvac::DetectorProperties::slideStartSize

stepX and stepY are respective to the nativeWindowSize and scaled along with the window To use default values for one parameter, use the value of -1; stopSize is inclusive. If the specified sizes do not have the same aspect ratio as the nativeWindowSize, the maximum range will be assumed. stepX and stepY are respective to the nativeWindowSize and scaled along with the window.

float cvac::DetectorProperties::slideStepX = MINUSONE

float cvac::DetectorProperties::slideStepY = MINUSONE

Size cvac::DetectorProperties::slideStopSize

int cvac::DetectorProperties::verbosity

The level of output verbosity from server to client; use -1 for default.

double cvac::DetectorProperties::videoFPS = -1.0

If applicable, process frames in a video at the specified number of frames per second. This fps is a suggestion to the algorithm which might or might not observe it. It is also understood as an approximate value since convertion is probably not feasible for any number of fps. Special values: -1: use the specific trainer/detector's default setting -2: process at the native video frame rate -3.n: process every n-th frame in the video

---

The documentation for this struct was generated from the following file:

• src/Services.ice