Embedded Image Analysis System Based on B-ANN

Autor: Laurentiu-Mihai Ionescu, Gheorghe Serban, Alin Gheorghita Mazare, Ioan Lita, Daniel Alexandru Visan, Adrian Ioan Lita
Rok vydání: 2020
Předmět:
Zdroj: 2020 43rd International Spring Seminar on Electronics Technology (ISSE).
Popis: An increasingly common requirement is related to image analysis and automatic extraction of events “in situ” - on the place where image is acquired - without the need to send the image remotely for analysis. There are integrated systems in the video cameras that allow the analysis of images especially in applications for perimeter monitoring and there is research done for different analysis systems that can be integrated near the camera. However, these existing solutions either do not allow high levels of analysis or, if they allow, involve the use of expensive and difficult to implement technologies. In this paper we present a solution, based on B-ANN that allows image analysis, classification and extraction of events of different types. The proposed solution can be implemented on a FPGA circuit for commercial use and allows, through the digital and analog interfaces, the connection to different types of cameras.
Databáze: OpenAIRE