Scholarlink

<p>Two-dimensional (2D) semiconductor-based photodetectors possess unique physical properties and flexible device architectures, showing tremendous potential for multidimensional optoelectronic information acquisition and intelligent perception. Compared with commercial photodetectors, 2D photodetectors can detect intensity, wavelength, polarization, and temporal information owing to the intrinsic properties of 2D materials. By integrating advanced device engineering with intelligent algorithms for processing multidimensional optical signals, 2D photodetectors enable information reconstruction within a single device, thereby advancing next-generation machine vision technologies. This review summarizes recent progress in machine vision based on 2D photodetectors. It first discusses the extension of sensing dimensions enabled by 2D material properties. It then examines device-level working mechanisms, including steady-state, synaptic, and spiking response modes. Next, it highlights algorithmic integration in applications such as spectrometry, image recognition, and motion detection. The review also discusses structural innovations for single-device optical field detection and outlines future directions in high-dimensional information fusion. By merging device engineering with computational intelligence, this work presents a framework for multidimensional perception and intelligent processing, and underscores the transformative role of 2D photodetectors in machine vision. Looking ahead, continued advances in device integration and algorithmic co-design are expected to enable efficient, real-time visual systems for complex environments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Advances in machine vision based on two-dimensional photodetectors: device engineering and computational intelligence

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月22日

查看全文
  • Jiaxin Li,
  • Xiaojie Tang,
  • Enen Zhang,
  • Lelun Shu,
  • Ziqi Zhou,
  • Fagang Jiang,
  • Xinghua Wang,
  • Yali Yu,
  • Zhongming Wei

摘要

Two-dimensional (2D) semiconductor-based photodetectors possess unique physical properties and flexible device architectures, showing tremendous potential for multidimensional optoelectronic information acquisition and intelligent perception. Compared with commercial photodetectors, 2D photodetectors can detect intensity, wavelength, polarization, and temporal information owing to the intrinsic properties of 2D materials. By integrating advanced device engineering with intelligent algorithms for processing multidimensional optical signals, 2D photodetectors enable information reconstruction within a single device, thereby advancing next-generation machine vision technologies. This review summarizes recent progress in machine vision based on 2D photodetectors. It first discusses the extension of sensing dimensions enabled by 2D material properties. It then examines device-level working mechanisms, including steady-state, synaptic, and spiking response modes. Next, it highlights algorithmic integration in applications such as spectrometry, image recognition, and motion detection. The review also discusses structural innovations for single-device optical field detection and outlines future directions in high-dimensional information fusion. By merging device engineering with computational intelligence, this work presents a framework for multidimensional perception and intelligent processing, and underscores the transformative role of 2D photodetectors in machine vision. Looking ahead, continued advances in device integration and algorithmic co-design are expected to enable efficient, real-time visual systems for complex environments.