论文部分内容阅读
本文以跟踪电视系统中自适应量化器为设计背景,提出了一种新的、实时自适应的快速图象量化方法——逐极均值法,文中首先用Lloyd-Max最佳量化理论分析了这种量化方法的均方误差失真,讨沦了图象中存在孤立亮点时的处理方法。然后论述了这种量化方法应用于跟踪电视系统中的性能,即实现的简单、快速性;对照度变化的自适应性;及图象对比度增强效果。文中通过图象处理实验结果验证了这种量化方法的性能和理论分析的正确性。最后得出结论:逐极均值法量化器是一种能够代替LlodyMax最佳量化器的次佳量化器,这种量化器可以很好地满足跟踪电视系统中对自适应量化器的设计所提出的各方面性能要求;它对那些要求实现简单、实时自适应的量化器应用领域也将具有一定意义。
In this dissertation, based on the adaptive quantizer in tracking TV system, this paper proposes a new fast real-time adaptive image quantization method, ie, the pole-by-average method. In this paper, we firstly analyze this by using the optimal quantization theory of Lloyd-Max The mean square error of the method of quantification distorts the treatment when there are isolated bright spots in the image. Then it discusses the application of this quantization method to tracking the performance of TV system, that is to say it is simple and fast to implement; the adaptiveness of the change of illumination; and the enhancement effect of image contrast. The results of image processing experiments verify the performance of this method and the correctness of theoretical analysis. Finally, it is concluded that the pole-by-average method quantizer is a sub-optimal quantizer that can replace the optimal quantizer of LlodyMax. The quantizer can well meet the requirements of the tracking TV system for adaptive quantizer design Performance requirements in all aspects; it will also be of some value for those applications of quantizers that require simple, real-time adaptation.