Hierarchical method for foreground detection using codebook model

Abstract

This study presents a new hierarchical scheme with coarse level and fine level for foreground detection using codebook model. The code book is mainly used to compress information to achieve high efficient processing speed. In the coarse level, six intensity values are employed to represent a block. The algorithm extends the concept of the Block Truncation Coding (BTC), and thus it can further improve processing efficiency. In detail, the coarse level is divided into two stages: Level one can increase processing speed and reduce noises without increasing False Positive (FP) rate; level two can increase detected precision of level one. Fine level can further enhance precision in coarse level. Moreover, this study also presents a new color model which can classify an input pixel as shadow, highlight, background, or foreground with the match function. This model can also cooperate with the Mixture of Gaussian (MOG) to remove shadow and thus enhances MOG's performance. As documented in the experimental results, the proposed algorithm can provide superior performance to that of the former Codebook (CB) approach.

Statistical feature bag based background subtraction for local change detection

“…The source code for DPGMM scheme is taken from 15 . The parameters A C C E P T E D M A N U S C R I P T considered are set as advised in 16 . Similarly, the source code for p-ROST is obtained fromhttps://github.com/FlorianSeidel/GOL.…”

“…Here the authors have used the concept of codebook for construction of the background model and hence does not need estimation or fixation of any parameter. Guo et al [16] proposed a hierarchical BGS scheme where both block and pixel based codebooks are explored for object detection. Recently, Guo et al [15] have proposed a multilayer codebookbased BGS scheme where the adaptive features from the blocks of various sizes were combined to model nonstationary background.…”

Skip Decision and Reference Frame Selection for Low-Complexity H.264/AVC Surveillance Video Coding

“…However, the computational cost required to determine the foreground probability model and the sampler map limited the speedup obtained (over [10]) to 3. Guo et al [31] used a hierarchical, block and pixel-level segmentation technique to reduce computational complexity. Unlike in GMM S-MD, spatial samplers are not adopted and hence every pixel in the frame is accessed to compute the block-level features.…”

Fast image moving object segmentation based on block texture for embedded system implementation

“…For obtaining the better segmentation quality in various background scene, many studies employ complex background modeling strategy such as MoG model [1] [2], codebook model [3] [4], and texture model [5] [6]. However, their memory consumption and computation effort fully depend on the scene complexity.…”