DCT变换被广泛采用在图象和视频压缩标准中(如JPEG,MPEG,H 261 H 263) 而对于这些压缩图象的处理,传统的手段是先解压到空间域,再进行处理和识别,因而增加了计算复杂性。针对这个问题,运用压缩域上图像处理技术,提出了人脸特征表达的方法,并构造DCT域上人脸识别系统。在人脸标准库FERET上的测试,提出的算法与经典的特征脸法和PCA+LDA法相比较,不仅提高了精度,而且降低了计算复杂
RAW不再只是专业人士的玩物:那些使用RAW格式拍摄照片的用户,现在有了更多的选择可以 让他们的照片获得最棒的效果。你还可以从生产商那里免费获得你所需要的软件。你了解RAW格式 吗?你在使用它吗?
The basin-and-range area in eastern North China is known for frequent occurrence of earthquakes, their great magnitudes and heavy losses thereby incurred. Seismic studies in the past usually emphasize
The borehole strain meter at Wushi station recorded obvious anomaly before Jiashi MS= 6.8 earthquake occurred on February 24, 2003. Its features are as follows. j Anomaly types are complete. The tren
Field experimental seismic sounding permitted us to obtain optimal shallow seismic reflection sounding parameters. In process of data processing, we obtained a high-qualitative shallow seismic reflec
According to the critical point hypothesis (CPH), energy release would accelerate in power law before occurrence of large earthquakes or failure of brittle materials. In the paper, CPH was studied by
In the paper, for the application of stochastic simulation of ground motion, we put forward a method to determine ″the combined effect of amplification and attenuation″ (combined effect for short) of
上期比较详细的示出读写机具和感应卡片在模拟前端的作用过程,两个谐振回路LR／CR、LT／CT必须工作在谐振频率点附近,例如fo=125kHz。读写器模拟前端工作在持续谐振方式,感应卡片以相似的电路谐振在该频率点上(或略偏高-因为这种邻近耦合方式对谐振回路加载较重,会强制性降低卡片的
Fingerprint image is a typical non-restraint image that has some uncertainty, which makes it difficult to perform identification using classical approach. Therefore, fuzzy pattern recognition is appli