| 基于差分进化鲸鱼优化算法的森林冠层图像分割 |
| |
| 引用本文： | 邬博文,朱良宽,王璟瑀.基于差分进化鲸鱼优化算法的森林冠层图像分割[J].西北林学院学报,2022,37(1):67-73. |
| |
| 作者姓名： | 邬博文 朱良宽 王璟瑀 |
| |
| 作者单位： | （东北林业大学 机电工程学院，黑龙江 哈尔滨 150040） |
| |
| 基金项目： | 国家自然科学基金(31370710);黑龙江省博士后启动基金(LBH-Q13007)。 |
| |
| 摘 要： | 针对森林冠层图像结构复杂，受光照不均匀，导致分割精度较差等问题，提出一种基于差分进化鲸鱼优化算法的冠层图像分割方法。首先选取多阈值Kapur熵作为适应度函数，应用鲸鱼优化算法（whale optimization algorithm,WOA）对阈值的搜索过程进行优化，代替传统遍历搜索机制，加强阈值搜索能力与效率。其次引入差分进化算法（differential evolution algorithm,DEA）进行混合，采用以种群平均适应度为评价指标的混合策略平衡2种算法，增强算法的局部搜索能力，实现更好地平衡探索与开发。最后通过对森林冠层图像进行阈值分割研究，在适应度值、PSNR、SSIM与计算时间上进行对比分析。结果表明，本算法可以获得更精确的分割阈值和更高的分割精度。
|
| 关 键 词： | 鲸鱼优化算法 差分进化算法 Kapur熵 图像分割 森林冠层 |
Forest Canopy Image Segmentation Based on Differential Evolution Whale Optimization Algorithm |
| |
| WU Bo-wen,ZHU Liang-kuan,WANG Jing-yu.Forest Canopy Image Segmentation Based on Differential Evolution Whale Optimization Algorithm[J].Journal of Northwest Forestry University,2022,37(1):67-73. |
| |
| Authors: | WU Bo-wen ZHU Liang-kuan WANG Jing-yu |
| |
| Institution: | (College of Mechanical and Electronic Engineering,Northeast Forestry University,Harbin 150040,Heilongjiang,China) |
| |
| Abstract: | In order to solve the problems of complex structure and uneven illumination of forest canopy image,which lead to poor segmentation accuracy,a canopy image segmentation method based on differential evolution whale optimization algorithm(WOA)was proposed.Firstly,multi-threshold Kapur entropy was selected as the fitness function,and the WOA was used to optimize the threshold search process to replace the traditional traversal search mechanism to enhance the threshold search ability and efficiency.Secondly,the differential evolution algorithm(DEA)was introduced to mix,and the hybrid strategy with the average fitness of the population as the evaluation index was used to balance the two algorithms,so as to enhance the local search ability of the algorithm and better balance the exploration and development.Finally,through the research on the threshold segmentation of the actual canopy image,the fitness value,PSNR,SSIM and computing time were compared and analyzed.The experimental results showed that the proposed algorithm could obtain more accurate segmentation threshold and higher segmentation accuracy. |
| |
| Keywords: | whale optimization algorithm differential evolution algorithm Kapur entropy image segmentation forest canopy |
| 本文献已被 维普 等数据库收录！ |
| 点击此处可从《西北林学院学报》浏览原始摘要信息 |
| 点击此处可从《西北林学院学报》下载免费的PDF全文 |
