Inner-Distance-Based Shape Recognition of Target Object Using Binary Sensors

摘要

We propose an algorithm for shape recognition of the target object using the notion of the inner distance. The inner distance of the object going across the field, where binary sensors are densely deployed, are obtained from the distance between two binary sensors concurrently detecting the object. To estimate the position of each binary sensor, we apply a range-free localization using a moving landmark, which wanders around the field of watch. We numerically study the accuracy of the shape recognition based on the inner distance when it is combinedly used with the range-free sensor localization based on the moving landmark.