Some applications require the identification of the type and position of various objects in a part of the space (such as outdoors or in a hall). Such functionalities can be used in so called “smart homes”. One concrete example could be a voice controlled environment sensitive light switch.
This thesis proposes a plan for the flow of signal processing and an implementation for various localization and source identification algorithms. The localization algorithms delay the incoming signal from the microphone in time or frequency domain. Summing these signals performance is calculated. The maximum of performance provides the supposed source position. The source identification is divided into two parts. First, the representation of sound is generated from time functions, in the form of feature vectors. Then, the feature vectors are classified, i.e. source is determined. Upon completion of measurements, results are evaluated.