Automated optimization and restoration for underwater images

Software image processing using objective image quality metric to provide a simple, easily comparable measurement of image quality

Photonics Software & Information Technology

The quality of images taken underwater is vital to many military and civilian applications involving mine detection, diver visibility, and search and rescue operations. Unlike in the atmosphere, where visibility can be on the order of miles, the visual range underwater is limited to, at best, tens of meters, even in the clearest waters. Hence, obtaining sharp images relies on image reconstruction techniques. Although traditional image enhancement can be applied underwater images, their effectiveness is considerably limited because they do not take into account the processes that lead to the degraded images, which involve in-depth understanding of the optical properties of the water – an understanding necessary to compensate for blur and improve images.

To provide such knowledge of the underwater optical environment, Navy researchers have developed the image quality metric (IQM) which is based on a weighted grayscale angle (WGSA) – a measure of the overall sharpness of edges in the image.

The method begins with evaluating the unprocessed image according to the IQM and from there establishing a baseline WGSA. Restoration then starts with a set of data representing estimated optical properties of the water. The estimated optical properties are used to obtain a restored or true image. The quality of the restored image is then assessed by applying the IQM to the image and calculating a revised WGSA to represent the quality of the restored image. The revised WGSA is then compared to the previously calculated WGSA to determine whether the WGSA can be improved by further optimization. If based on the optimization algorithm used, a determination is made that the WGSA can be improved, the restoration cycle begins again, using the second set of estimated optical properties. This process continues until the WGSA can no longer be improved and the best-restored image is output.

The use of an objective image quality metric permits the use of the automated restoration scheme of the invention since it enables the computer to know which direction to go in restoring an image in small improvement increments, and know when to stop.

While developed by the Navy for better imaging in seawater, the technology and method can be applied to any type of imaging in a medium that may have optical properties relating to scattering or attenuation, such as imaging in foggy, smoggy, or smoky conditions. It could also be applied in medical imaging, where the image may be affected by biological scattering media such as blood and tissues.

This US patent 8,983,222 is related to US patents 8,204,328; 8,437,569; 8,437,568; and 8,639,055.

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

Contact Us