Spectral-360: A Physics-Based Technique for Change Detection

Abstract

Abstract-This paper presents and assesses a novel physicsbased change detection technique, Spectral-360, which is based on the dichromatic color reflectance model. This approach, uses image formation models to computationally estimate, from the camera output, a consistent physics-based color descriptor of the spectral reflectance of surfaces visible in the image, and then to measure the similarity between the fullspectrum reflectance of the background and foreground pixels to segment the foreground from a static background. This method represents a new approach to change detection, using explicit hypotheses about the physics that create images. The assumptions which have been made are that diffuse-onlyreflection is applicable, and the existence of a dominant illuminant. The objective evaluation performed using the 'changedetection.net 2014' dataset shows that our Spectral-360 method outperforms most state-of-the-art methods.

CDnet 2014: An Expanded Change Detection Benchmark Dataset

“…Maddalena and Petrosino's SC-SOBS method [15] also uses machine learning but with a self-organizing neural map. Spectral-360 [16] is based on the correlation between the diffuse spectral reflectance components of a new video frame and an evolving background model derived from recent training frames. The method of [30] compares the current pixel value with one long-term and several short-term adaptive templates that are discarded based on a measure of efficacy rather than age or random selection.…”

Bio-inspired Boosting for Moving Objects Segmentation

“…The bio-inspired motion segmentation module is running with no configurable parameters. For evaluation purposes, several alternatives were used as BS method: MOG2, refers to the masks outputted by MOG2, available in OpenCV, using default parameters; GMM [7], KNN [7], AMBER [11], CwisarDH [12], Spectral360 [13], SuBSENSE [14] and FTSG [15] refer to the computed masks made available in the CDnet site [9]. These masks were generated with the parameters adjusted to maximize overall performance.…”

Physics-based and Retina-inspired Technique for Image Enhancement

“…Another important parameter is the set of geometrical features, which represent the scene structure, the illuminant orientation, the surface roughness and the viewing geometry. These features combine non-linearly to form a digital image (Sedky, 2014).…”

“…While the word physics refers to the extraction of intrinsic features about the materials contained in the scene based on an understanding of the underlying physics which govern the image formation. This process is achieved by applying physics-based image formation models that attempt to estimate or eliminate the illumination and/or the geometric parameters to extract information about the surface spectral reflectance (SSR) (Sedky, 2014).…”