Long wavelength infrared imaging for advanced facial recognition

System creates a 3-D model from thermal images on a pixel-by-pixel basis


These examples of conventional images captured by a thermal imaging camera illustrate the lack of contrast addressed by the invention.

Long wavelength infrared (LWIR) imaging is generally a passive tool because it is a measure of a signal emitted by an object, thus not requiring an interrogating light source. But one disadvantage of LWIR imaging is the lack of contrast relative to background objects and noise. Images of people tend to look washed out and ghostlike, and it is difficult to obtain information from the images, which would allow reliable and repeatable identification of a person’s unique facial or other features.

Army scientists and engineers have greatly improved upon LWIR to enhance identifiability of a person. This device and system generate a 3-D model of the surface of a face or any object emitting thermal radiation using an image captured by a thermal camera. The image contains several component parts of data for each pixel in the image. For each pixel, a surface plane is established by calculating a set of angles with reference to the camera’s line-of-sight and the base of the camera. Once the surface plane is established, a normal vector for each pixel is calculated. The normal vectors for all the pixels are integrated over the entire image, forming a function representing the surface of the object. If the result of the function at each (x, y) position is computationally-rendered or plotted against a third axis (z), the topology of the surface of the object can be visualized in three-dimensions.

Do you have questions or need more information on a specific technology? Let's talk.

Contact Us