News | Jun 6, 2018
15 technologies for digital imaging, machine vision enhancement
These advanced digital imaging technologies are available to businesses for integration into new or existing products and services
Seeing the world around them is critical in order for robots to think and react to dynamic environments. That means machine vision (MV), the fusion of artificial intelligence and CMOS image sensors, is at the core of the automation revolution.
“The MV camera market will grow from $2B in 2017 to roughly $4B in 2023,” according to a recent industry report.
From food inspection and sorting to the guidance of autonomous cars, an expanding list of imaging companies are now building products for the growing list of work being done by automated systems.
Manufacturers of CMOS image sensors like Canon, Sony, and OmniVision can leverage years of applied research performed in U.S. military laboratories.
For example, scientists at the U.S. Navy’s Space and Naval Warfare Systems Command (SSC Pacific) have invented a real-time image defogger for enhanced contrast.
The image defogging technology and 14 others are listed below with links to more information. Each is available to businesses for development or integration into products and services.
TechLink, owner of the only DoD-wide patent database, is ready to assist. Since 1999, TechLink’s staff have provided hundreds of companies across the U.S. with no-cost licensing assistance, facilitating the transfer to industry of over 1,000 DoD inventions.
Air Force researchers have developed a novel 2x4 pattern of microgrid polarizers that delivers a 24 percent spatial resolution improvement over the traditional 2x2 array. Their solution yields better image quality from a camera without affecting the underlying detector array. This major design advancement hinges upon maximization of spatial frequency bandwidth by minimizing the risk of aliasing – a better separation of Stokes spectral components over the 2x2 case.
Navy researchers have developed a system to address scenes that are affected by fog that need to be enhanced in near real-time for the purpose of video processing and computer vision applications. The contrast enhancement methodology is based on the physics model of the scene and obtains an arbitrary-to-scale image of scene depth and an enhanced image. The leap in speed derived by this method is achieved by employing separable filters that approximate the FIR Wiener filter for refining the image of scene depth estimates.
Related to the above technology, the Navy has developed a joint turbulence mitigation and fog removal method with near real-time performance. In this patented process, a single image is de-noised, defogged, aligned, and de-blurred. A turbulence mitigation step completes the process.
The U.S. Army’s patented algorithm effectively increases the resolution of the reconstructed output image using the existing imaging device from a sequence of low resolution, undersampled imagery. By removing aliasing due to undersampling, SRIR can also improve range performance of the sensor. *Express licensing available
Army scientists and engineers have greatly improved upon LWIR to enhance the 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.
For optimization of a multi-sensor system, it’s necessary to quantify the gain in performance (resolution) attributed to each sensor and the effective resolution of the combined sensor image. Air Force scientists have developed an imaging system to test the resolving capability of an array of sensors each being responsive to a different range of electromagnetic radiation. The Quad Emissive Display (QED) provides a resolution target with a high contrast image that can be sensed in all spectral bands simultaneously. Figure 1 shows a schematic of the arrangement of components that make up the QED. *Research paper available
Contained in a pair of patents obtained by the U.S. Air Force, this morphological approach is more computationally efficient and more accurate than other contemporary methods of assessing the relative improvement in image quality when an image-enhancing algorithm has been applied to a digital image.
Navy scientists have invented a way to improve the TSPI data of multiple high-speed cameras with a calibration mechanism for measuring with precision the timing of light integration for visible and infrared cameras. In general, the video-timing test equipment utilizes a single frame step camera or video recorder to measure the accuracy of video overlay and metadata time stamp. It further incorporates a GPS coordinate universal time message signal along with a one pulse per second output signal to generate a timing accuracy of approximately one microsecond. For shutter speeds up to 1,600 frames per second.
Critical to machine vision is image analysis that identifies changes in a scene. It becomes exponentially more difficult when trying to register and fuse time-varying image sources.
To deal with this issue, Air Force scientists have developed a unique method for registering and fusing time-varying image sources to provide the highest possible information rendering.
This Air Force invention has significant advantages over other image recognition routines. The system obtains visual images of various rotations containing an object of interest using multi-spectral sensors and represents those images in a matrix format. A CPU determines histograms for each image and averages them into a single image histogram.
The Navy has developed a miniature cryogenic shutter assembly that allows small- and medium-format infrared cameras to perform high-dynamic range or superframe imaging (combining thermal images taken at different temperature ranges) by providing an absolute “null” offset that could only previously be done by systems using more expensive and cumbersome external reference methods.
The patented invention, called array set addressing, or ASA, solves the problems for any hexagonally arranged data sampling elements. This technology has broad imaging applications because hexagonal sampling works in all frequency bands. Licensing this technology creates a potential for collaboration with world-class Department of Defense imaging experts.
Navy researchers have developed a two-band imaging system which uses a dual-chrome component for radiometric measurement of objects. This system is designed to take the place of two cameras and solve the issue of spatial and temporal registration arrays.
The Air Force has simplified the analysis of rich hyperspectral data. In this system, spectral and temporal data is received and formulated into a vector/matrix. At that point, feature extraction is performed by any of the common methods including independent component analysis, non-linear prime component analysis, support vector decomposition, or combinations of these.
Navy researchers have developed a unique system of multilayer coatings useful in window and dome applications, which often use lossy EMI shields to protect sensitive visible and infrared sensors.
Search TechLink’s database, which includes thousands of technologies available to your business for new product development.
TechLink, the Department of Defense’s national partnership intermediary, provides businesses with no-cost assistance when applying for a patent license agreement or cooperative research and development agreement with DoD laboratories.