<title>Data fusion through fuzzy reasoning applied to segmentation of multisensory images</title>

Abstract

Multi-sensor systems provide a purposeful description of the environment that a single sensor cannot offer. Fusing several types of data enhances the recognition capability of a robotic system and yields more meaningful information otherwise unavailable or difficult to acquire by a single sensory modality. Because observations provided by sensors are uncertain, incomplete, and/or imprecise, we adopted the use of the theory of fuzzy sets as a general framework to combine uncertain measurements. We developed a fusion formula based on the measure of fuzziness. This fusion formula satisfies several desirable properties. We established a fuzzification scheme by which different types of input data (images) are modeled. This process is essential in providing suitable predictions and explainations of a set of observations in a given environment. After fusion, a defuzzification scheme is carried out to recover crisp data from the combined fuzzy assessments. This approach was implemented and tested with real range and intensity images acquired using an Odetics Range Finder. The goal is to obtain better scene descriptions through a segmentation process of both images. Despite the low resolution of the images and the amount of noise present, the segmented output picture is suitable for recognition purposes.

Hierarchical overlapped SOM's for pattern classification

“…Further increase in the overlap did not offer any substantial improvements in the performance. We also experimented with different combinations of overlap in the training and testing data such as (11,5), (13, 7), and (13, 5). These combinations yielded performances lower than those presented in the table for the same degree of training sample overlaps (refer to the last two rows in the table).…”

Degree of familiarity ART2 in knowledge-based landmine detection

“…The implementation of the fuzzy rule based system in the preliminary test using a fusion equation [11] identified the three landmine targets (including surrogate) with three false alarms as shown in Fig. 4.…”

“…1 gives the equation for the familiarity flag familiarity flag = cos (qi 0 ; Ii) = hqi 0 ; Iii e + kqi 0 k kIik: (11) …”

“…Data fusion function used in[11]. landmine) to one (high likelihood); (2) the position and the number of pixels in an 8 2 8 sized pixel window produced by the processed polarization resolved image shown inFig.…”