RIMAC project: Building a low cost camera network for feeding an open urban routing information system

Abstract

Provision of reliable traffic data could cause a reduction of the traffic problem. In a previous paper, the authors developed the core of RIMAC, an information system that allows the introduction of new and reliable sources of information from camcorders, traffic lights or mobile phones. With these new data sources, RIMAC is able to provide many more realistic reports. This paper describes the development of the second part of the project: the interconnection of the core with a source of information: a camera network. In order to avoid congestion or loss of data packets, thinking of non-optical fiber networks, speed reports will be sent to the RIMAC core instead of unprocessed images. In situ, at each camera node, speed calculation is done using a C++ computer vision application and the OpenCV's Libraries over an SBC (Single Board Computer) Raspberry Pi2 with a CMOS (complementary metal oxide semiconductor) camera. The first testbed was done in a few streets of Lima, an emerging South American city where live about 10 million people. Lima is beginning a long-term transportation reform plan, activists are pushing for introducing the right paradigms of citizenship and appropriate technology, like RIMAC, into the plan.