Related Experiment Video
Updated: Aug 5, 2025
09:56
Automated 3D Optical Coherence Tomography to Elucidate Biofilm Morphogenesis Over Large Spatial Scales
Published on: August 21, 2019
7.0K
Plastic Classification Using Optical Parameter Features Measured with the TMF8801 Direct Time-of-Flight Depth Sensor
Cienna N Becker1, Lucas J Koerner1
1Department of Computer and Electrical Engineering, University of St. Thomas School of Engineering, St. Paul, MN 55105, USA.
Sensors (Basel, Switzerland)
|March 30, 2023
Summary
This study presents a novel method for non-contact plastic classification using a low-cost direct time-of-flight (ToF) sensor. The technique achieves high accuracy, paving the way for smart material identification in consumer electronics.
Area of Science:
- Optoelectronics
- Materials Science
- Sensor Technology
Background:
- Accurate material identification is crucial for recycling and product safety.
- Existing methods often require physical contact or complex equipment.
- Miniaturized sensors offer potential for on-device material analysis.
Purpose of the Study:
- To develop a non-contact method for classifying different plastic types.
- To utilize inexpensive direct time-of-flight (ToF) sensors for material characterization.
- To explore physics-based optical parameters for robust classification.
Main Methods:
- Employed a direct ToF sensor (AMS TMF8801) to capture light return times from five plastic types.
- Collected ToF histogram data across various sensor-to-material distances.
- Trained a classifier using ToF data and fitted data to a physics-based scattering model.
Main Results:
- Achieved 96% classification accuracy using raw ToF histogram data.
- Developed a physics-based model using optical parameters (surface/subsurface scattering ratio, distance, decay time).
- This model yielded 88% accuracy and provided insights into scattering mechanisms.
Conclusions:
- Direct ToF sensors can effectively classify plastic types non-contact.
- Physics-based optical parameters derived from ToF data offer robust material identification.
- The methodology is suitable for integration into consumer electronics like smartphones.
