The advantages of using millimeter-wave radars for material classification.

Abstract

Millimeter-wave (mmWave) radars allow for greater capabilities for material classification when compared to pre-existing models. Current models primarily utilize RGB cameras which depend solely on the appearance of an object. The mmWave radar offers higher accuracy in material detection because it relies on identifying radio frequencies rather than images. The mmWave radar can also be used in various environments - including those that are pitch-black or even filled with smoke - without compromising accuracy levels. On the other hand, the accuracy of RGB cameras is affected when used in similar conditions. This study focuses on material classification using mmWave radars in an indoor environment and it involves small-scale tests to evaluate the accuracy of mmWave radars when differentiating between four types of materials: plastic, wood, metal, and water. Our goal is to measure the intensity value of the 3D points correlating to each material and analyze how the value differs by object. We hypothesize that the intensity value will differ by material, also correlating to how much light is absorbed by each material type. We seek to use mmWave radars in conjunction with RGB cameras to improve existing models by allowing for greater accuracy in material classification under diverse environments.