Convolutional neural network for building exterior and interior classification (Kemerovo, Tomsk)

Abstract

Purpose: The concept development of convolutional neural network for real-time distinguishing between building interior and exterior.

The relevance of the topic of the scientific article is determined by the technology integration into everyday life. Houses are increasingly referred to as a smart house. One of the elements of this control system, is robot vacuum cleaner, which cleans various surfaces. Difficulties encountered by such a technique largely depend on the environment definition, in which it is located.

The intelligent system can independently distinguish between building interior and exterior, thereby greatly increasing the performance of the firmware complex of modern technology in both domestic and industrial segments.