A research team led by Academician Zhang Jun at Beijing Institute of Technology (BIT) has proposed, for the first time internationally, the original concept of spectral-dedicated computing and pioneered a novel on-chip spectral computing architecture, solving the long-standing challenge of online real-time analysis in hyperspectral imaging.

The research findings have been published in the journal Science under the title Hypervision: An On-Chip Hyperspectral Microsystem for Online Video-Rate Computational Imaging.

This milestone follows the team's earlier breakthrough in on-chip optical encoding published in Nature in 2024, which raised the light utilization efficiency of on-chip hyperspectral imaging from less than 25 percent to 74.8 percent.

Building on that foundation, the team has overcome a series of key technologies, including hardware acceleration for high-density data streams, fabrication of spectral-dedicated computing chips, and lightweight AI spectral reconstruction networks — completing a full engineering closed loop from theoretical exploration and chip development to complete-system fabrication.

The team independently developed HyperVision — the world’s first integrated online real-time hyperspectral imaging microsystem. The device integrates the self-developed spectral-dedicated computing chip HyperN and the broadband spectral imaging sensor HyperspecI, together with a high-capacity battery module and a high-definition display. Compact, lightweight, low-power consumption, and intelligent with long battery life, it operates without external power supply or remote computing clusters, stably delivering video-rate real-time hyperspectral imaging across the visible to near-infrared band.

HyperVision bridges the entire chain from laboratory theory to lightweight portable equipment and real-time online sensing applications, redefining the boundary of machine vision and opening new possibilities for strategic emerging fields such as aerospace remote sensing, autonomous driving, smart agriculture and precision medicine.

The team has long focused on frontier fields such as scientific detection in complex environments, establishing multiple research platforms in novel sensing devices and intelligent information processing, and pioneering the new research direction of on-chip optoelectronic sensing. The research has promoted the deep integration of integrated circuits, microelectronics, computer science, optical engineering and artificial intelligence, and injected solid original momentum into China's development in optoelectronic information.