Real-time crosstalk compensation in multizone Time-of-Flight sensors

Time-of-flight (ToF) sensors are essential for applications requiring real-time operation and precise distance measurements, including autonomous navigation, smart glasses, and advanced computer vision tasks. Recent ToF sensors utilize Single Photon Avalanche Diodes (SPADs) as receivers in combination with Vertical Cavity Surface Emitting Lasers (VCSELs) to achieve multi-zone ranging. However, crosstalk caused by reflections from different surfaces degrades measurement accuracy, reducing reliability in real-world conditions. This work presents a novel real-time crosstalk compensation method leveraging machine learning and signal processing techniques to adaptively distinguish true distance data from interference artifacts. Experimental results using the VL53L7CX sensor demonstrate significant improvements in measurement accuracy compared to industry-standard techniques.