NVIDIA DeepStream SDK API Reference

5.0 Release

 All Data Structures Files Functions Variables Typedefs Enumerations Enumerator Macros Groups Pages
NvDsInferDetectionParams Struct Reference

Detailed Description

Holds detection and bounding box grouping parameters.

Definition at line 186 of file nvdsinfer_context.h.

Data Fields

union {
   float   preClusterThreshold
 
}; 
 Holds the bounding box detection threshold to be applied prior to clustering operation. More...
 
float postClusterThreshold
 Hold the bounding box detection threshold to be applied post clustering operation. More...
 
float eps
 Holds the epsilon to control merging of overlapping boxes. More...
 
int minBoxes
 Holds the minimum number of boxes in a cluster to be considered an object during grouping using DBSCAN. More...
 
int groupThreshold
 Holds the minimum number boxes in a cluster to be considered an object during grouping using OpenCV groupRectangles. More...
 
float minScore
 Minimum score in a cluster for the cluster to be considered an object during grouping. More...
 
float nmsIOUThreshold
 IOU threshold to be used with NMS mode of clustering. More...
 
int topK
 Number of objects with objects to be filtered in the decensding order of probability. More...
 

Field Documentation

union { ... }

Holds the bounding box detection threshold to be applied prior to clustering operation.

float NvDsInferDetectionParams::eps

Holds the epsilon to control merging of overlapping boxes.

Refer to OpenCV groupRectangles and DBSCAN documentation for more information on epsilon.

Definition at line 201 of file nvdsinfer_context.h.

int NvDsInferDetectionParams::groupThreshold

Holds the minimum number boxes in a cluster to be considered an object during grouping using OpenCV groupRectangles.

Definition at line 207 of file nvdsinfer_context.h.

int NvDsInferDetectionParams::minBoxes

Holds the minimum number of boxes in a cluster to be considered an object during grouping using DBSCAN.

Definition at line 204 of file nvdsinfer_context.h.

float NvDsInferDetectionParams::minScore

Minimum score in a cluster for the cluster to be considered an object during grouping.

Different clustering may cause the algorithm to use different scores.

Definition at line 211 of file nvdsinfer_context.h.

float NvDsInferDetectionParams::nmsIOUThreshold

IOU threshold to be used with NMS mode of clustering.

Definition at line 213 of file nvdsinfer_context.h.

float NvDsInferDetectionParams::postClusterThreshold

Hold the bounding box detection threshold to be applied post clustering operation.

Definition at line 197 of file nvdsinfer_context.h.

float NvDsInferDetectionParams::preClusterThreshold

Definition at line 192 of file nvdsinfer_context.h.

int NvDsInferDetectionParams::topK

Number of objects with objects to be filtered in the decensding order of probability.

Definition at line 216 of file nvdsinfer_context.h.


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