Dodging Of UAV Aerial Images Based On The Mask Method

UAV aerial operations need to get the image post-processing, image uneven illumination phenomenon is a common and important problem, uneven illumination phenomenon UAV obtain aerial imagery showed uneven color and brightness distribution. Is well known, the light through the objective lens after the luminance on the focal plane is gradually decreased from the center to the surrounding, and also by the atmospheric attenuation, clouds concentration and the backlight and to the optical images of different degree. This requires aerial images acquired dodging a direct impact on the quality of the image after dodging processing the image quality and the amount of information feedback, and post-processing the image. This article is mainly for traditional Mask dodging processing method has been improved, the traditional Mask dodging processing is required to get a background image using the filter to select an appropriate scale, but this scale and can not fully use the images. Therefore, the basis of the Mask dodging law texture segmentation based on color image Mask dodging processing method. Technology and algorithms involved in the experiment, as follows is the research involved in technology research:1. Dodging processing direction research background, research purposes as well as domestic and foreign research status analysis caused by the image uneven illumination of external factors and internal factors, the research direction:Mask dodging processing based on color image texture segmentation method.2. Details of the conversion between several color model representation and color model, and research related to the texture analysis and research focuses on statistical methods inside the gray level co-occurrence matrix, the autocorrelation function, local binary mode and Laws texture energy measurement. Laid the foundation for the subsequent experiments.3. Mainly introduces the basics of the texture, detail the statistical methods inside the gray level co-occurrence matrix, the autocorrelation function of the local binary patterns and Laws texture energy measurement. Then introduced the structural and spectral methods as well as the texture description GLCM feature extraction algorithm flow and prove the validity of GLCM feature extraction in the experiment. Have a significant role in the theory of the experiment, mainly used in image texture segmentation, and achieve the ultimate goal of this paper is based on image texture segmentation Mask dodging processing method has a key role.4. Mainly introduces the Mask dodging, Wallis uniform light approach, the Retinex uniform light processing method, and the homomorphism filtering dodging processing method, image dodging this regard has been developed for decades, can not be just these types of dodging, knowledge of this chapter describes several classic commonly used dodging and understand the principles and processes of mainstream dodging, and carried out experiments and analyzes. Then Mask dodging processing method slightly improved, better results can be obtained in the actual image on the image light absorbed.5. This paper presents an improved Mask dodging processing method experimental proof of the improvement can be achieved, and the original image of the same traditional Mask dodging processing hair experiments, compared with the improved method, proved improved Mask dodging processing method is superior to the traditional approach of dodging. To achieve the desired effect and purpose.