| Summary: | M.Ing. === Atmospheric turbulence (also referred to as optical or heat Scintillation, or heat shimmer) is a particular problem encountered in video surveillance, especially over distances where the target object focused on is over lkm in the distance. Images obtained from video surveillance are commonly required to be of a high quality for object identification and classification. Atmospheric turbulence causes degradation in the image quality through the blurring and a warping of the image, making object identification difficult. Algorithms have and still are being developed to suppress the image turbulence in digital video footage and enhance detail. There is a lack of reliable comparisons among algorithms to provide research direction, methods for identification of the best algorithms for particular applications, identification of useful image processing techniques and a full understanding of the problem. This need and lack of comparisons among the algorithms and atmospheric turbulence degraded videos is identified through the problem identification chapter. A literature study is undertaken in which the source of atmospheric turbulence and models are identified, image processing techniques discussed, filtering of electromagnetic waves reviewed, a review of some equipment, and a discussion of metrics. This is followed by the presentation of a number of atmospheric turbulence suppression algorithms developed by other authors. After a discussion of the algorithm implementations, the experimental design is described for algorithm image quality and performance investigation as well as the effect of optical filters. Experimental results are presented and discussed which provide repeatable results pertaining to the algorithms' image quality and processing requirements. The results allowed identification of the algorithms' strengths and weaknesses, how they compare, and their suitability for real and post processing environments. Efficient performing software components were also able to be identified, particularly Illuminance-Reflectance adjustment. The experiments and results provide a solution to this atmospheric turbulence comparison problem.
|
|---|
