Black ice and aquaplaning make streets dangerously slippery and can cause severe accidents. Assistance systems that are able to detect such road conditions and respond to them – for example by automatically reducing speed – would significantly enhance road safety. On behalf of Intel (funded as part of the center of excellence “Securely Networked Systems”) Fraunhofer EMFT scientists are working on a visual real-time system that is capable of detecting water on a road surface in its solid state.
The catch here is that within the visible spectrum, it is difficult to distinguish water from other materials, depending on quantities and light conditions. In order to avoid this problem, the development team is using the infrared spectral range for visual detection. Based on the mutual relationship between the water molecule vibrations, researchers are able to detect the solid state of water on various surfaces. The absorption of light reflects the emission of the molecules’ natural vibration in different ways from one material to the next. This emission cannot be measured directly, however: it is calculated based on measurable variables (reflection or transmission).
In order to be able to offer the automotive industry an attractive solution, the researchers are testing to see whether low-cost CMOS image sensors as already used in smartphone cameras might also be suitable for detection. These sensors are mainly used for visible light but they are also sensitive to the near infrared range.