A Background Modeling and Foreground Detection Algorithm Using Scaling Coefficients Defined With a C
A Background Modeling and Foreground Detection Algorithm Using Scaling Coefficients Defined With a Color Model Called Lightness-Red-Green-Blue
This paper presents an algorithm for background modeling and foreground detection that uses scaling coefficients, which are defined with a new color model called lightness-redgreen- blue (LRGB). They are employed to compare two images by finding pixels with scaled lightness. Three backgrounds are used: 1) verified background with pixels that are considered as background; 2) testing background with pixels that are tested several times to check if they belong to the background; and 3) final background that is a combination of the testing and verified background (the testing background is used in places, where the verified background is not defined). If a testing background pixel matches pixels from previous frames (the match is tested using scaling coefficients), it is copied to the verified background, otherwise the pixel is set as the weighted average of the corresponding pixels of the last input images. After the background is computed, foreground objects are detected by using the scaling coefficients and additional criteria. The algorithm was evaluated using the SABS data set, Wallflower data set and a subset of the CDnet 2014 data set. The average F measure and sensitivity with the SABS Data set were 0.7109 and 0.8725, respectively. In the Wallflower data set, the total number of errors was 5280 and the total F-measure was 0.9089. In the CDnet 2014 data set, the F-measure for the baseline test case was 0.8887 and for the shadow test case was 0.8300.
Comparing an image with another image that was “scaled in lightness” by illumination changes, computing a background model and performing foreground detection. For the first problem, we define a new color model, called Lightness-Red-Green-Blue (LRGB), and scaling coefficients. These scaling coefficients can be used for comparing two images and for detecting illumination artifacts. For the last two problems, we propose a background model and foreground detection algorithm that uses the scaling coefficients. This algorithm achieves significant improvements in the SABS Dataset over the algorithms analyzed, and in the Wallflower Dataset over the algorithms analyzed.
If the current image has foreground objects, these objects may have pixel values that do not belong to the background, and the mapping may not be defined for these objects. The approach that is proposed in this work, finds if a current frame pixel value is a lightness scaled version of a background pixel value, and does not have this limitation.
Chromaticity coordinates indicate the portion of the color components in the intensity. This work proposes an alternative approach to describe the chromaticity and intensity using the LRGB color model.
This approach tries to find if a pixel is a scaled version of another pixel (all the components are scaled by the same constant) plus added noise. The method proposed in this work is different because it finds if the pixel is the lightness scaled version of other pixel, in the LRGB color space.