Enhanced VOT on Foggy Videos Using Probabilistic Estimation Scheme

Objectives: Visual object tracking allude the process for tracking the path of the object in the video sequence. Tracking of the object under the foggy environment is the challenging task in the video sequence. There is a great need to implement spectral analysis for the visual attention tracking model. The proposed methodology deals with object tracking using Dubbed Density of Fog Assessment based Defogger (DEFADE) filter to resist environment based detail restoration. Methods/Statistical analysis: DEFADE guided filter is used to improve the visual detail of the frame. Then frames are moved to learning network where particle region is detected and then is registered using Binary Pixel Distance Mapping (BPDM). Further Calculation of color values for the given object or selected object in the entire three channels RGB (Red, Green, Blue). The spectral analysis of visual attention tracking model is compared with the proposed approach using BPDM_CMF model. In this system an existing filter for binary conversion and color filtering was used to evaluate and estimate the trajectory of the tracked objects in foggy weather affected videos. In case of visual tracking and object detection, multi-scale detection methods constitute to combine the effort of object detection in spatial scale using BPDM with Color Mean Filter (CMF) and a defogging based spurious response inherent method for multiscale removal of fog in tracking videos and improve the fine edge tracking with reduce error probability detection between frames in a temporal tracking method. Findings: The execution of the proposed framework has been contrasted and the before technique. The proposed method works more efficiently under the foggy environment for tracking the objects in the video sequence. Application/Improvements: Automatic Recognition of the object in the video sequence. Defogging has been efficiently used in identifying the humans on the basis of trajectory