The Research On Image Denoising And Enhancement Of Panoramic Vision

In recent years, with rapid development of the large CCD array imaging technology and multi-DSP parallel processing technology, visual perception system is widely applied in many other fields as the basic three-dimensional space perception approach. Panoramic visual images are received by the lens-based imaging field of view larger than the hemisphere(360°×180°) visual information images, because of its open sight, for civilian, military, space and aerospace applications in the field of paramount importance.In this paper, we adopted the methords of fish-eye lens and mirror which have been widely concerned currently to obtain the panoramic visual images.The image is offen impacted by the noise of surrounding environment as the image collection and transmission process, resulting in a decline in quality, and in certain circumstances by the light brightness and other factors have led to lack of clarity, the research on image denoising and enhancement technology is particularly important.As the rich details of large panoramic visual image,the ordinary image denoising method is difficult to achieve while meeting a good denoising effect and a high computing efficiency, a new proved methord which is based on the combining of wavelet that has well time-frequency localized characteristics and fourth-order partial differential equations that dos well in retaining the inage edges and details, is proposed in this paper for image denoising.After the denoising of panoramic visual image, in order to get better visual the study based on partial differential equations (PDE) and pulse coupled neural network (PCNN) in image enhancement is focused researched in this paper. Gray-scale images from the entire stretch with the details of the distribution of the enhanced two-pronged, this paper presents the impact of partial differential equations based on the filter image model and the local histogram equalization method of combining pairs of gray-scale images processed and applied to color image processed. Finally for this algorithm, as the selection of some important parameters directly affecting the final picture of the effect of using particle swarm optimization algorithm to optimize the selection, to achieve optimal parameter.