近日，北京理工大学唐鑫团队研究了带有CMOS传感器集成上转换器的红外可视化蛇启发人工视觉系统。2025年8月20日，《光：科学与应用》杂志发表了这一成果。

生物视觉激发了具有多种视觉功能特征的人工视觉系统的发展，然而，检测到的波长仅在0.4和0.78μm，限制了其应用。蛇有探测和转换红外线的窝状器官，因此可以产生动物的热成像，即使在黑暗中也能准确地瞄准捕食者或猎物。

受自然红外视觉蛇的启发，研究组提出了利用CMOS传感器集成上转换器的人工视觉系统，以打破可见光限制，首次实现3840 × 2160超高分辨率短波红外（SWIR）和中波红外（MWIR）可视化成像。通过胶体量子点势垒异质结结构的红外探测单元设计以及共基质发光单元的引入，该器件在室温下实现了短波红外区6388.09 cd m⁻²的发光亮度和6.41%的上转换效率，中波红外区1311.64 cd m⁻²的发光亮度和4.06%的上转换效率。该人工视觉系统拓宽了红外领域的广泛应用，如夜视、农业科学和工业检测，标志着生物人工视觉的重大进步。

附：英文原文

Title: Infrared visualized snakes-inspired artificial vision systems with CMOS sensors-integrated upconverters

Author: Mu, Ge, Lin, Yangye, Fu, Kerui, Tang, Xin

Issue&Volume: 2025-08-20

Abstract: Biological visions have inspired the development of artificial vision systems with diverse visual functional traits, however, the detected wavelength is only in visible light between 0.4 and 0.78μm, restricting their applications. Snakes generate a thermal image of animals due to pit organs for detecting and converting infrared, allowing them to accurately target predators or prey even under darkness. Inspired by natural infrared visualized snakes, we propose artificial vision systems with CMOS sensors-integrated upconverters to break visible light limitations to realize 3840×2160 ultra-high-resolution short-wave infrared (SWIR) and mid-wave infrared (MWIR) visualization imaging for the first time. Through colloidal quantum dot barrier heterojunction architecture design of infrared detecting units and the introduction of co-hosted emitting units, the luminance and upconversion efficiency reach up to 6388.09cdm2 and 6.41% for SWIR, 1311.64cdm2 and 4.06% for MWIR at room temperature. Our artificial vision systems broaden a wide spectrum of applications within infrared, such as night vision, agricultural science, and industry inspection, marking a significant advance in bioartificial vision.

DOI: 10.1038/s41377-025-02001-x

Source: https://www.nature.com/articles/s41377-025-02001-x