基于连通区域的仿射不变区域提取方法?

针对现有仿射不变特征提取方法存在计算量大、鲁棒性差的问题,提出一种基于连通区域的仿射不变区域提取方法。对输入的灰度图像进行各向异性高斯滤波,并对滤波后的图像进行灰度直方图均衡化。找到图像中灰度值相同点所组成的连通区域,将灰度值差小于delta的相邻连通区域进行合并,把满足条件的最后一次合并结果作为图像的局部仿射不变区域。实验证明,该方法提取效果好、速度快且鲁棒性强。     

连通区域提取算法研究

常见的连通区域提取算法是针对四连通区域的。本文分析了几个相关的算法，并指出其优劣点，提出了一个八连通区域提取算法。首先按照一定的标号规则扫描图像，得到一定数量的连通区域；然后针对同一区域内标号不一致的情况进行处理即得到最终处理结果。实验结果表明此方法能有效地提取八连通区域，故可用于连通区域复杂的图像如签名的分块处理中。     

采用特征跟踪的裂缝缺陷三维重建

通过对连续切片图像中裂缝区域跟踪实现裂缝缺陷的三维重建。分为以下三个步骤：区域预测、区域变形和表面绘制。预测是根据当前切片图像区域对下一个切片图像的区域进行预估。由于预测的结果和实际区域不一致采用区域变形技术得到实际的区域。对得到的区域进行表面绘制完成三维重建。在连续的切片图像中对提取出的区域不需对应关系的计算且在交互过程中利于特征的选取,因此计算成本低。实验结果表明通过上述方法可以对单个裂缝区域进行快速的跟踪和重建。     

基于图像处理的人造板孔穴含量统计的新方法

针对人造板横切面孔穴含量的微观图像的特点,经常要对一幅图像中多个连通区域的面积进行计算。因此,提出了一种通过一次扫描图像,即能快速求出图像中多个连通区域的面积的新方法,并通过实验对两种封闭区域面积统计方法进行比较。结果表明在处理连通区域较多的面积统计时,新方法计算速度较快。     

基于区域生长法提取二值图像中的连通区域

标记二值图像连通区域是图像处理过程的基本算法,机器视觉和模式识别中常用此方法提取目标和分析目标几何特征。文章以人机交互方式获得初始种子点,增强种子点的可靠性,通过区域生长法提取二值图像中的连通区域。     

基于乳腺钼靶X线图像等值面特征的乳腺区域阈值分割算法

乳腺钼靶X线图像中乳腺区域的分割可以帮助对图像进行深入分析和处理,从而提高乳腺疾病的诊断准确率。提出一种能有效提取乳腺区域的算法。算法分析了乳腺钼靶X线图像等值面面积变化不连续的特征并将其用于分割阈值的精确计算。该算法使用基于扫描线的方法来获得含乳腺区域的连通区域,比种子填充法效率更高。为了获得更纯粹的乳腺区域,通过一些精细地处理对乳腺区域相连的未曝光图像边框作了剥离。实验结果表明算法在乳腺区域分割的精度和执行效率上都有更好的表现。     

基于图像分层树的图像语义分割方法

针对传统语义分割模型缺乏空间结构信息,无法准确地描述对象轮廓的问题,提出了一种基于图像分层树的图像语义分割方法。分层树模型采用结构森林方法生成轮廓模型,为防止过度分割,运用超度量轮廓图算法得到多尺度轮廓图,然后利用支持向量机训练多尺度轮廓图生成图像分层树,通过随机森林精炼分层树,最终输出图像语义分割结果。在测试实验中,像素精确度达到82.1%,相比区域选择方法(Selecting Regions)提升了2.7%。并在较难区分的树和山脉的预测精确度上,相比层次标记方法(Stacked Labeling)分别提升了16%,25%,具有更高的稳定性。实验结果表明,在复杂的室外环境下,对图像语义分割的精确度、稳定性和速率均有明显改善。     

图像中多语种文本提取的高斯混合建模方法

建立了相邻字符区域的高斯混合模型,用于区分字符与非字符.在此基础上,提出了一种从图像中提取多语种文本的方法.首先对输入图像进行二值化,并执行形态学闭运算,使二值图像中每个字符成为一个单独的连通成分.然后根据各连通成分重心的Voronoi区域,形成连通成分之间的邻接关系;最后在贝叶斯框架下,基于相邻字符区域的高斯混合模型计算相应的伪概率,以此为判据将每个连通成分标注为字符或非字符.利用所提出的文本提取方法,进行了复杂中英文文本的提取实验,获得大于97%的准确率和大于80%的召回率,证实了方法的有效性.     

基于特征模糊推理的形态学颗粒分割算法

针对粘连或重叠颗粒图像的分割问题,提出了一种基于特征模糊推理的局部形态学重构参数计算方法,对传统的距离变换结合分水岭的算法进行了改进。在传统距离变换结合分水岭方法的基础上,将颗粒图像划分成若干连通区域,每个连通区域单独处理,使用形态学局部重构的方法抑制分水岭的过分割现象。通过对距离图像连通区域极大值进行统计分析,提取该连通区域的颗粒形态特征。将颗粒形态特征作为模糊输入,重构参数特征作为模糊输出,使用模糊推理方法自适应地计算重构参数,解决了重构参数选取的不确定性问题。最后对重构图像进行分水岭变换得到颗粒分割图像。实验结果表明,该方法对各种粘连状态的颗粒分割效果良好,克服了传统方法的过分割与参数自适应选择的问题。     

提取轮廓图背景控制点的立体匹配算法

针对动态规划匹配算法的误匹配与狭窄遮挡物问题,提出一种在轮廓图中提取背景控制点的立体匹配算法,在立体图像对的轮廓图中选择背景控制点,利用动态规划在视差空间图像中搜索最优路径,根据视差约束以及狭窄遮挡物的判定公式完善视差图。仿真实验结果表明,该算法能够降低在视差不连续区域匹配上的误匹配率。     

结合全卷积网络的无监督视频目标分割

对视频中的目标进行像素级分割是计算机视觉领域的研究热点，完全没有用户标注的无监督视频分割对分割算法提出了更高的要求。近几年在分割中常使用基于帧间运动信息进行建模的方法，即用光流等运动信息预测目标轮廓，再结合颜色等特征建立模型进行分割。针对这些方法产生的前景背景混淆以及边缘粗糙等问题，本文提出结合全卷积网络的视频目标分割方法。首先通过全卷积网络预测视频序列中显著目标的轮廓，结合光流获得的运动显著性标签进行修正，然后建立时间-空间图模型，运用图割的方法获得最终的预测标签。在SegTrack v2以及DAVIS这2个通用数据集上进行评估，结果表明本文方法较基于帧间运动信息的方法在分割效果上有明显的提高。     

Pitch Detection Enhancement Employing Music Prediction

Pitch detection methods are widely used for extracting musical data from digital signals. A review of those methods is presented in the paper. Since musical signals may contain noise and distortion, detection results can be erroneous. In this paper a new method employing music prediction to support pitch determination is introduced. This method was developed in order to override disadvantages of standard pitch detection algorithms. The new approach utilizes signal segmentation and pitch prediction based on musical knowledge extraction employing artificial neural networks. Signal segmentation allows for estimating the pitch for a single note as a whole, therefore suppressing errors in transient and decay phases. Pitch prediction helps correcting pitch estimation errors by tracking musical context of the analyzed signal. As it was shown in the experimental results, pitch estimation errors may be reduced by using both signal segmentation and music prediction techniques.     

活动轮廓模型在重叠藻细胞计数中的应用

针对重叠藻细胞显微图像的计数问题,提出一种基于活动轮廓模型的重叠藻细胞自动计数方法。对重叠藻细胞显微图像进行预处理,标记连通域,提取并处理局部图像,采用活动轮廓模型算法提取重叠藻细胞边缘,同时进行边缘分析和计数。与传统分水岭算法相比,该计数方法不直接进行重叠藻细胞分割,可以消除因直接分割时过分割与欠分割导致的计数误差,提高重叠藻细胞计算准确度。实验结果表明,该方法对重叠藻细胞的计数准确率高于90%。     

基于感兴趣窄带区域的同步分割与配准方法及在IGRT中的应用

医学图像分割与配准是图像引导放疗(Image guided radiation therapy, IGRT)系统中的关键技术. 为提高基于CBCT (Cone beam CT)的IGRT系统实施胸腹部肿瘤放疗的实时性与自适应性, 特别是实现重要危及器官肝脏区域照射剂量的合理控制, 本文提出一种基于感兴趣窄带区域的同步分割与配准方法, 目标是实现放疗计划系统中计划CT和CBCT图像目标区域的分割与配准. 通过构建感兴趣窄带模型, 并且与活动轮廓模型相结合实现初始分割, 然后与基于光流场(Optical flow field, OFF)的形变配准方法进行循环迭代, 从而构造ASOR分割与配准同步模型(Active contour segmentation and optical flow registration synchronously, ASOR). 在方法实施时, 首先利用非线性扩散模型和窄带活动轮廓模型在CT图像中提取肝脏空间初始位置信息, 为同步模型提供合理的肝脏初始轮廓. 然后将该轮廓及相应窄带区域经仿射变换映射到CBCT图像, 进而结合构造的ASOR同步模型, 用光流场确定活动轮廓水平集的运动情况, 使分割与配准在同一个演化过程中完成迭代. 实验结果和临床应用表明, 本文提出的方法应用于基于CBCT的IGRT系统时, 可实现肝脏组织的自动分割与放疗剂量分布的快速计算. 同时, 我们将同步过程中获得的形变域用于实现肝脏与肿瘤靶区等剂量线从计划CT到CBCT的自适应转移, 进行自适应放疗效果的临床测评.     

多声部乐谱音符基元抽取方法

面向复杂版面的多声部乐谱,提出了符干、符头和符梁三类音符基元的抽取方法。提出基于垂直游程编码的粗提取、基于水平游程编码的精检测的符干抽取方法,能够有效克服基元间密集相交、粘连的干扰;在音符先验知识引导下,提出先分割、后特征检测的符头抽取方法,解决了粘连符头的切分难题;提出一种基于块状体分割和特征检测的符梁抽取方法,避开了传统的直线抽取方法所无法处理的符梁粘连难题。实验结果表明,该方法能够有效抽取复杂环境下多声部乐谱的音符基元。     

Rediscover flowers structurally

Existing methods for flower classification are usually focused on segmentation of the foreground, followed by extraction of features. After extracting the features from the foreground, global pooling is performed for final classification. Although this pipeline can be applied to many recognition tasks, however, these approaches have not explored structural cues of the flowers due to the large variation in their appearances. In this paper, we argue that structural cues are essential for flower recognition. We present a novel approach that explores structural cues to extract features. The proposed method encodes the structure of flowers into the final feature vectors for classification by operating on salient regions, which is robust to appearance variations. In our framework, we first segment the flower accurately by refining the existing segmentation method, and then we generate local features using our approach. We combine our local feature with global-pooled features for classification. Evaluations on the Oxford Flower dataset shows that by introducing the structural cues and locally pooling of some off-the-shelf features, our method outperforms the state-of-the-arts which employ specific designed features and metric learning.     

Multiple scale neural architecture for enhancing regions in the colour image segmentation process

Abstract: A dynamic multi‐scale neural model for enhancing regions and extracting contours in the colour image segmentation process is proposed. This model combines colour and textural information to coherently enhance images through the operation of two main components: the colour opponent system (COS) and the chromatic segmentation system (CSS). First, the COS module transforms the RGB chromatic input signals into long‐ (L), middle‐ (M) and short‐ (S) wavelength cone activations and luminance signals and then it generates the luminance (black–white), L–M and S–(L+M) opponent channels. The CSS module incorporates contour extraction, double opponency mechanisms and diffusion processes to yield coherent enhancing regions in colour image segmentation. This enhancement allows the region labelling to be more efficient due to a reduction in the uncertainty of the images' region allocation. The CSS module is based on the boundary contour system/feature contour system, extending it to allow colour stimuli processing to obtain general purpose architecture for image segmentation with later applications in computer vision and object recognition. Simulations show the good visual results obtained and the robustness of the architecture when processing images presenting different levels of noise. A benchmark using the Berkeley Segmentation Dataset is included in order to quantify and compare the results obtained.     

A robust region-based active contour for object segmentation in heterogeneous case

In this paper, we propose a new approach using a local version of the region-based active contour for object segmentation in images presenting heterogeneity in both the object of interest and the background. The local version was recently developed to deal with heterogeneous appearances by relying on extracting the local instead of global image statistics where the local extracted area is defined by using a disk, sketched at each point along the active contour, with a constant radius size inside and outside the contour. However, the use of a constant radius may prevent the active contour getting more information from its neighborhood and thus being trapped by undesired boundaries. To avoid this error segmentation, the local extracted area using our proposed approach is determined based on using two different radii to extract separately the interior and the exterior local information of the active contour. Using synthetic and real images, our approach shows an improvement in term of computation time and outperforms the conventional methods in noisy images presenting heterogeneity in both the object of interest and the background and using inadequate contour initialization.     

Contour evolution scheme for variational image segmentation and smoothing

An algorithm, based on the Mumford-Shah (M-S) functional, for image contour segmentation and object smoothing in the presence of noise is proposed. However, in the proposed algorithm, contour length minimisation is not required and it is demonstrated that the M-S functional without contour length minimisation becomes an edge detector. Optimisation of this nonlinear functional is based on the method of calculus of variations, which is implemented by using the level set method. Fourier and Legendre's series are also employed to improve the segmentation performance of the proposed algorithm. The segmentation results clearly demonstrate the effectiveness of the proposed approach for images with low signal-to-noise ratios.     

Watersnakes: energy-driven watershed segmentation

The watershed algorithm from mathematical morphology is powerful for segmentation. However, it does not allow incorporation of a priori information as segmentation methods that are based on energy minimization. In particular, there is no control of the smoothness of the segmentation result. In this paper, we show how to represent watershed segmentation as an energy minimization problem using the distance-based definition of the watershed line. A priori considerations about smoothness can then be imposed by adding the contour length to the energy function. This leads to a new segmentation method called watersnakes, integrating the strengths of watershed segmentation and energy based segmentation. Experimental results show that, when the original watershed segmentation has noisy boundaries or wrong limbs attached to the object of interest, the proposed method overcomes those drawbacks and yields a better segmentation.