中文摘要：

针对当前小目标图像阈值分割研究工作面临的难题,提出了快速核密 度估计图像阈值分割新方法.首先给出了基于加权核密度估计器的概率计算模 型,通过引入二阶Renyi熵作为阈值选取准则,提出了基于核密度估计的图像阈 值分割算法 （Kernel density estimator based image thresholding algorithm, KDET）, 然后通过引入快速压缩集密度估计 （Fast reduced set density estimator, FRSDE）技术,得到核密度估计的 稀疏权系数表示形式,提出快速核密度估计图像阈值分割算法fastKDET,并从 理论上对相关性质进行了深入探讨.实验表明,本文算法对小目标图像 阈值分割问题具有更广泛的适应性,并且对参数变化不敏感.

英文摘要：

In order to threshold the image containing small targets well, a novel fast kernel density estimator based image thresholding algorithm is proposed. Firstly, a novel computation model for probability density estimation based on the kernel density estimator with weighting coefficients is presented. By introducing the 2nd-order Renyi entropy as the threshold selection criterion, a novel kernel density estimator based image thresholding algorithm （KDET） is proposed. Then a fast version for KDET, named fastKDET, is proposed by integrating fast reduced set density estimator （FRSDE） and RSDE into the data condensation procedure. Moreover, some fundamental theoretical properties are fully studied. At last, several experiments are conducted and show that our fastKDET is more general than some of the existing algorithms and is insensitive to the parameters.