Development of a Hybrid Ultraviolet Imaging Algorithm for Optical Sensing Systems

This paper presents an undergraduate research experience in the design of a computer vision system by developing a novel combinational Ultraviolet (UV) image processing algorithm. The reflected-UV and UV fluorescence imaging methods are used in various scientific, industrial and medical optical sensing systems, such as in digital forensics, industrial fault inspection, astronomy, dermatology (monitoring skin conditions), germicidal technology and in remote sensing. UV electromagnetic spectrum is defined as a wavelength range from 10 nm (below is x-ray wavelength) to 400 nm (above is visible to human eye). Digital imaging in this broad range of wavelengths requires multiple optical lenses with efficient UV transmission (pass filters). A system's cameras, optics, filtering and illumination must be carefully selected according to the UV ranges being imaged. Most of the system design depends on custom-built cameras with modified lenses and UV filters that are relatively expensive to operate, maintain and repair. In addition, UV imaging generated by a camera and lens combination are device dependent. For instance, in reflected-UV imaging, UV illumination reflects of an object and is recorded by a UV-sensitive camera. UV fluorescence imaging is based on the UV illumination that stimulates fluorescence at a longer wavelength than UV excitation source. The resulting fluorescence and image are typically in the visible band and can be captured by a color camera. These optical sensing system specific results require high definition cameras with multispectral sensitivities. Thus, it is critical to provide an integrated and efficient approach to address the variability of UV based optical sensing systems. The objective of the proposed research is to develop a new adaptive UV sensitive image processing algorithm to transform our ability to combine reflected-UV and UV fluorescence techniques. The proposed algorithm uses a hyperspectral imaging technique to obtain the electromagnetic spectrum information from the pixels in the UV image to identify the wavelength range. The acquired data allows the system to adapt to the spectral range and to provide efficiency in the UV imaging methods within the system. The MathWorks' MATLAB software is being used to develop the proposed algorithm.