基于视觉注意的图像感兴趣区域分割算法
Segmentation algorithm of interest detection regions based on visual attention mechanism
-
摘要: 为了更好地对浅景图像和简单背景图像进行分析和处理,进而改善图像的质量和效果,在Itti注意力模型算法的基础上,提出了基于视觉注意的阈值分割算法.该算法主要通过对像素属性的分析提取图像的感兴趣部分,根据像素点与均值的偏离程度对图像进行增强,利用感知学原理使HSI颜色空间所成图像更符合人的视觉要求.实验结果表明,该方法符合生物学的视觉注意机制,在自动检测感兴趣区域时,可以有效减少过分分割,也能较好地提取出较大的感兴趣区域.
-
关键词:
- Itti注意力模型算法 /
- 特征显著性 /
- 图像感兴趣区域 /
- 图像分割
Abstract: In order to analyze and dispose shallow view images and simple background image preferably and improve image quality and effect,on the basis of Itti attention algorithm,the test method was proposed based on visual attention.Through the analysis of image pixels,this method extracted the interested part of the image,enhanced the image using the deviation degree between pixels and the average of image,and made image more accord with visual requirements using perception study principle.Experimental results showed that the proposed method complies with biological visual attention mechanism,which is also effective to reduce over-segmentation in auto-detecting ROI(region of interesting) and performs well for large objects.
计量
- PDF下载量: 18
- 文章访问数: 647
- 引证文献数: 0