The interest for ADL (Activity Daily Living) applications within smart homes, such as elderly monitoring, is increasing nowadays. Those applications are usually performed using machine learning algorithms that link actions with human activities which often depends on the human’s position. Human indoor localization is usually performed using Pyroelectric Infrared PIR sensors network, however, the latter presents several limitations, and to cope with these last, its results are sometimes combined with other sensing modalities to enhance location accuracy. In this paper, we present an idea to combine PIR network results with a non-intrusive Infra-red imaging sensor, relying on interval analysis IA techniques. We also show the advantages of using Ibex software to deal with IA calculations. Our future work is to characterize the error from the imaging sensor so that to encompass imaging result(s) within box(es) that could be combined, using Ibex, with PIR network box(es) so that to enhance the location accuracy.