云计算模型下图像边缘重叠区域检测方法研究
|
摘要
在对单目视觉图像检测中容易受到边缘轮廓区域分割误差的扰动影响,导致出现边缘重叠区域,需要对图像重叠区域进行准确检测,提高对单目视觉三维重建图像的自动识别和检测能力,提出一种基于多视角高分辨率重建和融合降噪的单目视觉三维重建图像边缘重叠区域检测技术.采用分块区域分割技术进行单目视觉三维重建图像的空间位置信息分布式识别,为边缘重叠区域检测提供点云数据,在云计算模式下采用局部空间融合降噪方法进行图像降噪,采用活动轮廓检测方法进行单目视觉三维重建图像的空置区域特征点定位,在边缘轮廓区域图像的分块区域内进行边缘重叠区域的动态特征提取,构建灰度直方图,结合边缘重叠区域动态特征的云检测技术,实现对图像边缘重叠区域检测.仿真结果表明,采用该方法进行图像边缘重叠区域检测的准确性较高,图像融合性较好,提高了重叠区域的图像识别能力.
In the detection of a monocular vision image, it is easy to be disturbed by the segmentation error of edge contour region, which leads to the appearance of overlapping edge area. Therefore it is necessary to accurately detect the overlapping region of the image. In order to improve the ability of automatic recognition and detection of monocular3 D reconstructed images, a new edge overlapping region detection technique based on multi-view high-resolution reconstruction and fusion de-noising is proposed. The spatial location information of monocular 3 D reconstruction image is distributed by using block region segmentation technology, which provides point cloud data for edge overlapping area detection. In the cloud computing mode, a local spatial fusion denoising method is used to reduce image noise. The active contour detection method is used to locate the empty region feature points of the monocular3 D reconstruction image, and the dynamic feature extraction of the overlapping edge region is carried out in the block region of the edge contour region image, and the gray level histogram is constructed. Combined with cloud detection technology of dynamic feature of edge overlapped region, the edge overlap region detection of image is realized.Simulation results show that the proposed method is more accurate in image edge overlapping region detection and image fusion is better, and it improves the ability of image recognition.
In the detection of a monocular vision image, it is easy to be disturbed by the segmentation error of edge contour region, which leads to the appearance of overlapping edge area. Therefore it is necessary to accurately detect the overlapping region of the image. In order to improve the ability of automatic recognition and detection of monocular3 D reconstructed images, a new edge overlapping region detection technique based on multi-view high-resolution reconstruction and fusion de-noising is proposed. The spatial location information of monocular 3 D reconstruction image is distributed by using block region segmentation technology, which provides point cloud data for edge overlapping area detection. In the cloud computing mode, a local spatial fusion denoising method is used to reduce image noise. The active contour detection method is used to locate the empty region feature points of the monocular3 D reconstruction image, and the dynamic feature extraction of the overlapping edge region is carried out in the block region of the edge contour region image, and the gray level histogram is constructed. Combined with cloud detection technology of dynamic feature of edge overlapped region, the edge overlap region detection of image is realized.Simulation results show that the proposed method is more accurate in image edge overlapping region detection and image fusion is better, and it improves the ability of image recognition.
引文
[1]浦瀚,杨道业,温勇.雾天景区图像增强的算法优化[J].中国仪器仪表,2018(10):66-68.
[2]LIM B,SON S,KIM H,et al.,Enhanced deep residual networks for single image super-resolution[C]//CVPRW 2017:Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops. Washington,DC:IEEE Computer Society,2017:1132-1140.
[3]YANG L. Multi-focus image fusion method based on NSST andⅡCM[C]//Proceedings of the 2017 International Conference on Emerging Internetworking,Data and Web Technologies. Berlin:Springer,2017:679-689.
[4]赵金波,肖照,白本督,等.基于L*a*b*颜色空间的高动态范围成像算法[J].计算机工程,2018,44(12):247-250,257.
[5]NIU S,CHEN Q,SISTERNES L D,et al.Robust noise region-based active contour model via local similarity factor for image segmentation[J]. Pattern Recognition,2016,61:104-119.
[6]余淮,杨文.一种无人机航拍影像快速特征提取与匹配算法[J].电子与信息学报,2016,38(3):509-516.
[7]SUDEEP P V,PALANISAMY P,RAJAN J,et al.Speckle reduction in medical ultrasound images using an unbiased non-local means method[J].Biomedical Signal Processing and Control,2016,28(6):1-8.
[8]赵蓉,顾国华,杨蔚.基于偏振成像的可见光图像增强[J].激光技术,2016,40(2):227-231.
[9]PRADHAN S,PATRA D. RMI based non-rigid image registration using BF-QPSO optimization and P-spline[J].International Journal of Electronics and Communications,2015,69(3):609-621.
[10]向文,张灵,陈云华,等.结合结构自相似性和卷积网络的单幅图像超分辨率[J].计算机应用,2018,38(3):854-858.
[2]LIM B,SON S,KIM H,et al.,Enhanced deep residual networks for single image super-resolution[C]//CVPRW 2017:Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops. Washington,DC:IEEE Computer Society,2017:1132-1140.
[3]YANG L. Multi-focus image fusion method based on NSST andⅡCM[C]//Proceedings of the 2017 International Conference on Emerging Internetworking,Data and Web Technologies. Berlin:Springer,2017:679-689.
[4]赵金波,肖照,白本督,等.基于L*a*b*颜色空间的高动态范围成像算法[J].计算机工程,2018,44(12):247-250,257.
[5]NIU S,CHEN Q,SISTERNES L D,et al.Robust noise region-based active contour model via local similarity factor for image segmentation[J]. Pattern Recognition,2016,61:104-119.
[6]余淮,杨文.一种无人机航拍影像快速特征提取与匹配算法[J].电子与信息学报,2016,38(3):509-516.
[7]SUDEEP P V,PALANISAMY P,RAJAN J,et al.Speckle reduction in medical ultrasound images using an unbiased non-local means method[J].Biomedical Signal Processing and Control,2016,28(6):1-8.
[8]赵蓉,顾国华,杨蔚.基于偏振成像的可见光图像增强[J].激光技术,2016,40(2):227-231.
[9]PRADHAN S,PATRA D. RMI based non-rigid image registration using BF-QPSO optimization and P-spline[J].International Journal of Electronics and Communications,2015,69(3):609-621.
[10]向文,张灵,陈云华,等.结合结构自相似性和卷积网络的单幅图像超分辨率[J].计算机应用,2018,38(3):854-858.