Time-compressive computational CMOS image sensors can realize a new implementation of light detection and ranging (LiDAR), i.e., pseudo-direct time-of-flight (pseudo-dToF) depth imaging. They are based on the multi-tap charge modulator typically used in indirect ToF (iToF) image sensors and compressive sensing of high-speed optical temporal signals in the charge domain at each pixel. Therefore, large digital circuits for time-to-digital conversion or histogram builder required in conventional dToF image sensors based on the single photon avalanche diode (SPAD) are not necessary. Nevertheless, they can detect multiple reflections under multipath interference, which is one of the advantages of dToF. Exploiting the small pixel size of the pseudo-dToF image sensor, high-resolution image sensors will be possible. Experimental results for a dual-reflection scene demonstrate the benefit of the pseudo-dToF.

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

Pseudo-Direct LiDAR Based on an Ultra-High-Speed Charge-Domain Compressive Sensing Image Sensor

  • Keiichiro Kagawa

摘要

Time-compressive computational CMOS image sensors can realize a new implementation of light detection and ranging (LiDAR), i.e., pseudo-direct time-of-flight (pseudo-dToF) depth imaging. They are based on the multi-tap charge modulator typically used in indirect ToF (iToF) image sensors and compressive sensing of high-speed optical temporal signals in the charge domain at each pixel. Therefore, large digital circuits for time-to-digital conversion or histogram builder required in conventional dToF image sensors based on the single photon avalanche diode (SPAD) are not necessary. Nevertheless, they can detect multiple reflections under multipath interference, which is one of the advantages of dToF. Exploiting the small pixel size of the pseudo-dToF image sensor, high-resolution image sensors will be possible. Experimental results for a dual-reflection scene demonstrate the benefit of the pseudo-dToF.