In this paper, a low complexity demodulation scheme is proposed for IEEE 802.15.4 LR-WPAN systems. In the proposed scheme, multiple active correlators for demodulation are replaced with a matched filter-based cross-correlator shared with a synchronization unit, which requires only a few additional control units. Consequently, with the proposed demodulation scheme, the total hardware complexity for the LR-WPAN baseband modem is 54% less than that of the existing architecture to have the separated demodulation and synchronization units.
Optical quadrature amplitude modulation (QAM) has attracted considerable attention because of its high spectral efficiency. For high capacity, the modulation rate should also be enhanced, which requires the generation of high-baud-rate multilevel electric signals with high precision in amplitude. In this paper, we describe a method to reduce the amplitude fluctuation of the electric signal by adopting a return-to-zero (RZ) format. We also demonstrate 2.0-Gbaud optical QAM signal generation with a conventional dual-parallel Mach-Zehnder modulator (DPMZM) driven by the electric RZ data signals.
In this paper, we propose an efficient shot change detection algorithm for MPEG compressed video by using a novel histogram feature essentially different from conventional ones. This algorithm is based on the adjacent pixel intensity difference quantization (APIDQ) histogram method, which had been reliably applied to human face recognition previously. The proposed algorithm has been evaluated by publicly reliable MPEG video database. Experimental results show the APIDQ histogram method can detect shot changes in MPEG video more accurately and robustly than conventional methods.
We propose new chaos-based communication systems in which on-line ICA algorithm is employed for signal separation in MIMO type receiver. The performance is evaluated in terms of the bit error rate (BER) under different levels of noise power. We derived the relation between BER and cross-talk error (CTE) in our systems. The BER is proportional to fourth root of the CTE in different two systems.
The human face localization task upon upright vertical frontal views faces in complex scenes is formulated as a wavelet-based problem and developed a novel approach using the extrema density aims to determine a single face position. Face-of-interest region is firstly located and framed by finding facial edges using the inter-orientation wavelet subbands and the anthropometric measure. Then, a refinement using the proposed head corner detection approach is carried to further improve the localization accuracy. Comparisons with existing state-of-the-art face detection works, showing that our system has a comparable performance in terms of face localization rate and quality.