搜索资源列表
View-Face-Detection
- 中科院的多视角人脸识别 Real-Time Multi-View Face Detection
Perspective-accurate-classification-of-face-detect
- 精确分类的视角无关人脸检测方法与硬件加速体系结构Perspective accurate classification of face detection method has nothing to do with hardware acceleration architecture-Perspective accurate classification of face detection method has nothing to do with hardware acceleration a
face
- 。基于颜色的系统可以在不同的视角中均能检测到人脸，不受人脸姿态变化的影响。根据颜色进行人脸检测的问题在于人所感知到的颜色与到达视网膜的光线的频率并不是一样的。通过对人脸图像做二值化，统计像素比例信息。得到人脸识别结果.这里通过对人脸图像打上网格，对区域块图像做二值分析，通过像素比例来做处理。进而得到人脸区域。-. Color-based system could be detected in a different perspective to the face, not the face at
siftDemoV4
- SIFT算法是近几年才提出的一种新型局部特征描述子。SIFT特征独特性好，信息量丰富，对尺度缩放，旋转，视角变化，遮挡，噪声和亮度变化等大部分干扰都具有很好的鲁棒性，它在在场景匹配，目标识别等领域已获得成功利用-SIFT algorithm is a new feature in recent years that it has made a partial descr iption of the child. SIFT features unique and good, rich amount
2015-research-frontier-Report
- 研究前沿的分析提供了一个独特的视角来 揭示科学研究的脉络。研究前沿的分析不依赖于对文 献的人工标引和分类-Analysis of Frontier provides a unique perspective to reveal the context of scientific research. Analysis does not depend on the forefront of the literature manual indexing and classification