Description: 用openCV编写的SURF特征检测方法，此代码只供学习交流，版权归原作者所有，如有商业利用请于原作者liuliu联系。-SURF prepared by openCV feature detection methods, this code is for learning exchanges, and belongs to original author, if the commercial exploitation of liuliu Please contact original author. Platform: |
Size: 6144 |
Author:于子予 |
Hits:
Description: A very good paper that explain tracking facial feature under varying face pose and facial expression-This paper presents a hierarchical multi-state pose-dependent approach for facial feature detection and tracking under varying facial expression and face pose. For effective and efficient representation of feature points, a hybrid representation that integrates Gabor wavelets and gray-level profiles is proposed. To model the spatial relations among feature points, a hierarchical statistical face shape model is proposed to characterize both the global shape of human face and the local structural details of each facial component. Furthermore, multi-state local shape models are introduced to deal with shape variations of some facial components under different facial expressions. During detection and tracking, both facial component states and feature point positions, constrained by the hierarchical face shape model, are dynamically estimated using a switching hypothesized measurements (SHM) model. Experimental results demonstrate that the proposed method accurately and robustly tracks faci Platform: |
Size: 958464 |
Author:Ng Jack |
Hits:
Description: In this thesis we present an operational computer vision system for real-time detection and tracking of human motion. The system captures monocular video of a scene and identifies those moving objects which are characteristically human. This serves as both a proof-ofconcept and a verification of other existing algorithms for human motion detection. An approach to statistical modeling of motion developed by Y. Song is coupled with a preprocessing stage of image segmentation and point feature tracking. This design allows a system that is robust with respect to occlusion, clutter, and extraneous motion. The results of experiments with the system indicate the ability to minimize both false detections and missed detections. Platform: |
Size: 778240 |
Author:linuszhao |
Hits:
Description: TLD(Tracking-Learning-Detection)是英国萨里大学的一个捷克籍博士生在其攻读博士学位期间提出的一种新的单目标长时间（long term tracking）跟踪算法。该算法与传统跟踪算法的显著区别在于将传统的跟踪算法和传统的检测算法相结合来解决被跟踪目标在被跟踪过程中发生的形变、部分遮挡等问题。同时，通过一种改进的在线学习机制不断更新跟踪模块的“显著特征点”和检测模块的目标模型及相关参数，从而使得跟踪效果更加稳定、鲁棒、可靠。-TLD (Tracking- Learning- Detection) is a Czech doctoral student at the university of surrey in the PhD put forward a new kind of single objective during long time (long term Tracking) Tracking algorithm. The algorithm with the conventional tracking algorithm of significant difference between the traditional tracking algorithm and the traditional detection algorithm to solve the combination of being tracked target deformation happened in the process of being tracked, partial sheltering and other issues. At the same time, through an improved online learning mechanism constantly updated tracking module "remarkable feature point" and a detection module of the target model and related parameters, so as to make the tracking performance is more stable, robust and reliable.
Platform: |
Size: 737280 |
Author:rambolyb |
Hits:
Description: ：基于背景建模的运动目标分割是智能视频监控的重要任务，模型的质量直接影响到检测、跟踪、识别等运动分析的准确性．当前的建模方法多是单层的，忽略了像素特征在时域和空域上的联系，模型描述不够准确，对于背景扰动、全局光照变化及复杂的室内外场景等多种情况鲁棒性不强，导致了分割中出现空洞和噪声点．针对这些问题提出了一种双层建模的方法，在第一层提取时域上的像素亮度特征采用码本建模，第二层提取邻域纹理特征采用基于中心对称的局部二值模式建模．实验证明该方法在用于运动分割时，比常用方法具有更好的准确性和鲁棒性．-Moving object segmentation based on background modeling is an important task of intelligent video surveillance, quality models directly affects the detection, tracking, recognition accuracy motion analysis. The current modeling methods are mostly single, ignoring the characteristics of pixels on the temporal and spatial linkages, model description is not accurate enough, the background disturbances, global illumination changes and complex indoor and outdoor scenes and other conditions are not robust strong, leading to a hollow-point division and noise appear. To solve these problems presents a double-modeling method, the pixel brightness feature extraction using time domain modeling codebook in the first layer, the second layer of texture feature extraction using neighborhood center symmetric local binary pattern based construction mode. Experiments show that the method used for motion segmentation when compared with the conventional method has better accuracy and robustness.
Platform: |
Size: 700416 |
Author: |
Hits:
Description: 用matlab实验运动图像的运动轨迹检测和特征点跟踪-Matlab experiment with the trajectory of motion detection and image feature point tracking Platform: |
Size: 4228096 |
Author:贺荧娇 |
Hits:
Search - detection and tracking of point feature