论文部分内容阅读
基于像斑的变化向量分析法CVA(Change Vector Analysis)过分依赖像斑的灰度均值信息,而未能有效利用其灰度分布信息,这在高分辨率遥感影像变化检测中存在不足。本文提出了一种基于像斑直方图相似性测度的变化检测方法。利用G统计量构建不同时期像斑之间的相似性测度。假设所有像斑的相似性测度值符合混合高斯分布模型,通过期望最大化算法EM(Expectation Maximization)求解相关参数,最后采用基于最小错误率的贝叶斯判别规则获取最终的变化结果。实验表明,本文提出的上述方法能够有效提高变化检测的精度。
Change vector analysis (CVA) based on speckle (CVA) over-relies on the gray-level mean information of image patches and fails to effectively use its gray-level distribution information. This is a disadvantage in high-resolution remote sensing image change detection. This paper presents a change detection method based on similarity measure of histogram. Using G Statistics to Construct Similarity Measurements Between Speckles at Different Stages. Assuming that the speckle similarity measures conform to the mixed Gaussian distribution model, the relevant parameters are solved by the Expectation Maximization (EM) algorithm, and finally the Bayesian discriminant rule based on the minimum error rate is used to obtain the final change result. Experiments show that the above method proposed in this paper can effectively improve the accuracy of change detection.