Various types of turbulences affect the visual quality during video surveillance. Turbulence such as fog or haze is responsible for image contrast. Removing this turbulence is a challenging problem in video surveillance. We used a method for stabilizing this atmospheric turbulence using dual tree complex wavelet fusion. Informative ROIs are selected using frame selection method from good quality frames. Morphological erosion is done afterwards. Image fusion is done using dual tree complex wavelet transform so as to fuse two images. In this paper, dark channel prior mechanism is proposed for haze removal. Finally image quality assessment is done. This proposed method is shown to outperform existing one and provide enhance visual quality in the range of surveillance scenarios. General TermsDual Tree Complex Wavelet Transform (DT-CWT).

Many valuable data in a video can be lost because of atmospheric distortion such as mirage, haze, fog, etc. This is a challenge in video surveillance system. We propose complex wavelet fusion for atmospheric turbulance algorithm for this purpose. This algorithm consists of series of steps to extract the information from the video. Frame selection method is used for selection of informative ROI only from good quality frame. These ROI were then fused to get the required objects. The fusion is performed in the Dual Tree Complex Wavelet Transform (DT-CWT) which employs two different real discrete wavelet transforms (DWT) to provide the real and imaginary parts of the CWT. Finally Haze removal is done by using dark channel prior mechanism. The proposed method is shown to significantly enhance the visual quality in a range of video surveillance. General TermsDual tree complex wavelet transform (DT-CWT).