Examples
Basic Examples (2) 
Filter out rectangles from list that have a big overlap with other rectangles in the list:
Define and display a list of rectangle shapes:
Delete rectangles with high overlap and display the results in red along with the original rectangles in black:
Only keep non-overlapping rectangles:
Scope (4) 
Perform non-max suppression on a list of regions:
Provide an explicit list of weights:
Performing non-maximum suppression eliminated some of the rectangles:
Show the result:
Ask for a specific property:
The index is determined by the order of the scores for rectangles that contained in the result:
Ask for a list of properties:
Ask for all properties:
Options (6) 
AcceptanceThreshold (2) 
Specify an AcceptanceThreshold for the regions:
Specify values to use before and after the overlap filtering:
MaxOverlapFraction (2) 
Use MaxOverlapFraction→f to only keep regions with an overlap smaller than f:
If soft NMS is used, MaxOverlapFraction is used to defined the scaling behaviour:
If the a scaling value is given explicitly, the option is ignored:
Method (2) 
By default, regions are removed using a hard overlap threshold:
Filter boxes using a soft NMS that scales the weights based on the overlap:
Specify the scaling value used to update the regions weights:
Soft NMS can become very slow if boxes are not removed:
Even a small threshold can significantly speed-up the computation:
Version History
-
1.0.1
– 06 December 2023
-
1.0.0
– 14 December 2021