摘要
在三维立体成像技术研究中,图像深度提取不受光源照射方向及物体表面发射特性的影响,不存在阴影,因此可以更准确地表现物体目标表面的三维深度信息。为了获取精确的深度图像,提出一种深度获取方法,采用基于改进的遗传算法的最佳熵阈值图像分割法对图像进行分割处理,进而得到深度图像。此方法可明显提高所得到的深度图像的准确度和有效性。同时,改进遗传算法可以快速逼近最佳阈值,大大缩短最佳阈值图像分割过程中阈值的选取时间,提高分割效率和深度获取的准确性。
In the research of three-dimensional imaging technology, the extraction of image depth is not affected by the illumination direction of the light source and the launch characteristics of the object surface, meanwhile with no shadow, and it can accurately show the three-dimensional depth information of the object surface. We propose a depth acquisition method to obtain the high-quality depth images, using the optimal entropy threshold image segmentation method based on improved genetic algorithm to segment the image and then obtain the depth image. This method can obviously improve the accuracy and real validity of the obtained depth image. The improved genetic algorithm can quickly approximate the optimal threshold, greatly shorten the threshold time of the optimal threshold image segmentation, and improve the accuracy of segmentation efficiency and depth acquisition.
作者
王琦
朴燕
Wang Qi;Piao Yan(College of Electronic and Information Engineering, Changchun University of Science and Technology Changchun, Jilin 130022, China)
出处
《激光与光电子学进展》
CSCD
北大核心
2018年第2期170-176,共7页
Laser & Optoelectronics Progress
基金
国家自然科学基金(60977011)
关键词
图像处理
三维立体成像
深度获取
熵
阈值分割
遗传算法
image processing
three-dimensional imaging
depth acquisition
entropy
threshold segmentation
genetic algorithm
