High-Resolution High-Sensitivity Human-Computer Algorithm and Gesture Control SOC Chip for Artificial Intelligence Applications

摘要

This paper presents a human-computer algorithm and gesture control SOC chip. Based on single CIS and active infrared method, the gesture control SOC chip realizes high-sensitive and high-precision discrimination of various human-computer interaction gestures by utilizing two main algorithms: target pattern extraction and specific gesture recognition. The SOC chip integrates all functional modules is implemented in 55 nm low power process with chip area of 4148*5148 um² and power consumption of 88 mW, providing advantages such as wide field of view of 600*600*600 mm³, high precision of 1 mm and high resolution 1 mm/pixel, which indicate that the presented highly integrated SOC chip has a great potential for artificial intelligence (AI) human-computer applications.