Improved Illumination Independent Moving Object Detection Algorithm Applied to Infrared Video Sequences

Improved Illumination Independent Moving Object Detection Algorithm Applied to Infrared Video Sequences

Copyright: © 2014 |Pages: 6
DOI: 10.4018/978-1-4666-4896-8.ch006
OnDemand PDF Download:
No Current Special Offers


Performance of the moving objects detection algorithm on infrared videos is discussed. The algorithm consists of two phases: the noise suppression filter based on spatiotemporal blocks including dimensionality reduction technique for a compact vector representation of each block and the illumination changes resistant moving object detection algorithm that tracks the moving objects. The proposed method is evaluated on monochrome and multispectral IR videos.
Chapter Preview

2. Improved Noise Resistant Illumination Independent Moving Object Detection Algorithm

The technique for moving object detection consists of two major phases:

  • 1.

    Image filtering of a current frame with the noise removal filter coefficients extracted with the PCA analysis.

  • 2.

    Detection of moving objects applying the pixel based method for moving object detection resistant to illumination changes.

A given video is treated as three-dimensional (3D) array of gray pixels with two spatial dimensions X, Y and one temporal dimension Z. We use spatiotemporal (3D) blocks represented by N-dimensional vectors, where a block spans (2T+1) frames and contains NBLOCK pixels in each spatial direction per frame N = (2T+1)× NBLOCK× NBLOCK, which can be found in the literature mentioned in the references. To represent the block vector by a scalar while preserving information to the maximal possible extent, principal component analysis is used. For principal component analysis, sample mean and covariance matrix of representative sample of block vectors corresponding to the considered types of movies are estimated and the first eigenvector of the covariance matrix S (corresponding to the largest eigenvalue) is used that represents the coefficients of the 3D filter that suppresses the noise. The 3D filter can be emulated by three 2D filters applied on frames z-1, z and z+1.

Complete Chapter List

Search this Book: