Video captured by night vision cameras often produces grainy and obscure thermal images, making it difficult to discern details. However, a new technique utilizing artificial intelligence (AI) has been developed to analyze thermal images, resulting in images that are nearly as clear as those taken in daylight with conventional cameras. This advancement has the potential to greatly enhance the night-time vision of self-driving cars and other technologies. While conventional cameras capture visible light, radar and lidar use radio waves and laser beams to map surroundings. However, these methods encounter difficulties in capturing images in the absence of reflected light and are susceptible to interference. Thermal imaging, which captures the infrared light emitted by objects, can be a valuable tool, but it has been underutilized in computer vision. One of the challenges is that every object emits heat signals, causing conventional thermal imaging to be plagued by excessive noise and clutter. To address this problem, a team of researchers has employed AI to train a neural network to identify an object's heat signature and separate it from the surrounding environmental noise, which creates the ghosting effect.

By recognizing the unique heat emission spectra of known materials, the algorithm characterizes the objects and removes them from the scene. This allows the algorithm to analyze the noise signals and reconstruct a detailed image with improved clarity and depth information. The team has named the technology "heat-assisted detection and ranging" (HADAR). Unlike radar and lidar that require signal emissions for data collection, HADAR is a passive technology. The researchers believe that HADAR could have a wide range of applications such as increasing the safety of self-driving cars and aiding wildlife tracking. Although the technology is promising, further work is required to overcome obstacles like motion blur and faster data collection. Despite these challenges, the researchers are optimistic about the potential of this technology in various real-world applications.