Search

Fuzzy C-means (FCM) algorithm is sensitive to image noise and only considers the image numeric feature information, ignoring the spatial constraint relationship between neighbor pixels. In addition, the membership degrees can't describe the image uncertainties effectively, which makes the FCM-based segmentation results inaccurate. Even though the advanced FCM algorithms incorporating local information are robust to image noise to some extent, the image details are not preserved well and the tiny regions are hardly segmented. To resolve this issue, this paper proposed an improved FCM segmentation method based on intuitionistic fuzzy set (IFS), namely IFS_FCM. The proposed method integrated IFS into FCM, which well considered the image uncertainty. Furthermore, it added the spatial neighbor information into the objective function, which improved the robustness to image noise and preserved the image details. The experimental results demonstrate that the image segmentation results of this method are satisfactory. [ABSTRACT FROM AUTHOR]

To solve the problem of signal separation in the system of integration of radar and communication, a signal model based on random stepped-frequency Orthogonal Frequency Division Multiplexing ( OFDM) radar pulses is proposed. By the frequency agility, the communication data is modulated on the radar pulses and then transmitted through random stepped- frequency pulses. Thus, the target range image can be acquired from the correlation of the echo signal and the reference signal, while the communication data is demodulated by means of frequency detection through bandpass filter group without signal separation. Finally, simulation results show that the designed signal realizes decimeter range resolution imaging and high-rate data transmitting in Mbit per second, which can satisfy the requirements of transmitting the mass data. [ABSTRACT FROM AUTHOR]