#!/usr/bin/env python
from __future__ import division
import cv2
import numpy as np
from FrameProcessor import FrameProcessor
'''
Still in progress....
Filter to remove the background and only keep the parts of the video that are
actively changing.
'''
class BackgroundRemove( FrameProcessor ):
def __init__ ( self ):
super( BackgroundRemove, self ).__init__()
self._name = "Background Remove"
self._speed = 0.01
self._avg = None
self._fgbg = cv2.BackgroundSubtractorMOG()
def prop_AdaptSpeed_set( self, val ):
self._speed = float( val / 100 )
print( self._speed )
def prop_AdaptSpeed_get( self):
return int( self._speed * 100 )
def type_AdaptSpeed( self ):
return int
#def prop_AlgorithmV_get( self ):
# pass
def processFrame( self, frame_in ):
# version 1 - moving average
if self._avg == None:
self._avg = np.float32( frame_in )
cv2.accumulateWeighted( frame_in, self._avg, self._speed )
background = cv2.convertScaleAbs( self._avg )
active_area = cv2.absdiff( frame_in, background )
#version 2 - MOG - Gausian Mixture-based Background/Foreground Segmentation Algorithm
fgmask = self._fgbg.apply( frame_in ,learningRate = 0.01 )
#active_area = cv2.bitwise_and( frame_in, frame_in, mask = fgmask )
return fgmask
