Clutter modeling in infrared images using genetic programming

Abstract

Background clutter characterization in infrared imagery has become an actively researched field, and several clutter models have been reported. These models attempt to evaluate the target detection and recognition probabilities that are characteristic of a certain scene when specific target and human visual perception features are known. The prior knowledge assumed and required by these models is a severe limitation. Furthermore, the attempt to model subjective and intricate mechanisms such as human perception with general mathematical formulas is controversial. In this paper, we introduce the idea of adaptive models that are dynamically derived from a set of examples by a supervised learning mechanism based on genetic programming foundations. A set of characteristic scene and target features with a demonstrated influence on the human visual perception mechanism is first extracted from the original images. Then, the correlations between these features and detection performance results obtained by visual observer tests on the same set of images are captured into models by a learning algorithm. The effectiveness of the adaptive modeling principle is discussed in the final part of the paper.

Display Clutter

“…Overall, the r (correlation coefficient) and R 2 (coefficient of determination) values are relatively large, suggesting that the algorithms are good predictors of the performance effects of clutter. Some methods stand out with particularly large correlation coefficients, such as the approach of Voicu et al (2000), a genetic algorithm-based technique that requires a training phase in order for the algorithm to learn to recognize clutter. More recent techniques, such as the one by and , also result in R 2 values greater than .8.…”

“…The disagreements regarding what constitutes display clutter stems, to some extent, from the fact that the phenomenon has been studied and defined in many different domains. Furthermore, in fields such as radar research and electro-optical imaging, there is a tendency not to define clutter at all, with the implicit assumption that the meaning of clutter is clear to everyone (e.g., Chen, Tharmarasa, Kirubarajan, & Pelletier, 2011;Grahame, Laberge, & Scialfa, 2004;Klick, Blumenau, & Theriault, 2001;Roy & Kumar, 2010;Voicu, Uddin, Myler, Gallagher, & Schuler, 2000;Xu et al, 2009;Yang, Wu, Li, & Zhang, 2011). In the following section, we will review and discuss the various definitions of clutter to date.…”

A new approach of visual clutter analysis for pedestrian detection

“…Wolfe [16] proposed to use features including contrast, orientation, color and motion to measure clutter. Voicu et al [12] proposed a clutter model to measure infrared images. Global features and local features are computed and applied to train a genetic model to classify the clutter level.…”

“…Therefore, those proposed models which are well correlated with the systematically controlled experiments may not be directly suitable for real life situations. In addition, some of those models are tested and applied on clutter measure from a specific category, such as geospatial displays [10], infrared images [12]. The naturalistic driving scenes have very different characteristics that are associated with pedestrian appearance perception difficulty.…”