Image fusion is becoming one of the hottest techniques in image processing. Its performance evaluation method that can compare and analyze different fusion techniques is an essential part of image fusion techniques. In this paper, we proposed two intuitive schemes—correlation and information deviation schemes—for evaluating the performances of image fusion techniques. The former scheme is a fast and compact version of the quantitative correlation analysis (QCA). The latter scheme is an information deviation analysis. Using the two schemes, the performances of different fusion techniques can be compared directly and quantitatively based on the source images and the fused images with a faster speed than the QCA method. Two multiresolution analysis-based image fusion methods, wavelet transform-based fusion and pyramid transform-based fusion, which operate on typical hyperspectral image sets [airborne visible/infrared imaging spectrometer (AVIRIS)], are evaluated by the proposed schemes. The simulation results show the correctness and effectiveness of the proposed two schemes.

This paper presents a wavelet transform-based data fusion algorithm for multisensor systems.With this algorithm, the optimum estimate of a measurand can be obtained in terms of minimum mean square error (MMSE). The variance of the optimum estimate is not only smaller than that of each observation sequence but also smaller than the arithmetic average estimate. To implement this algorithm, the variance of each observation sequence is estimated using the wavelet transform, and the optimum weighting factor to each observation is obtained accordingly. Since the variance of each observation sequence is estimated only from its most recent data of a predetermined length, the algorithm is selfadaptive. This algorithm is applicable to both static and dynamic systems including time-invariant and time-varying processes. The effectiveness of the algorithm is demonstrated using a piecewisesmooth signal and an actual time-varying flow signal.

In order to develop an intelligent flow sensor which can be validated by itself, a wavelet-transform-based new method for assessing blockage of the sensing line in a differential-pressure flow sensor is proposed in this paper. The wavelet transform, a powerful means of signal processing for analysing the characteristic of a signal if there is little prior knowledge about it, is used to decompose the sensor-output signal into some elementary building blocks. The energy distributions in these building blocks reveal the effects of blockage of the sensing line of the sensor and then are exploited to determine their existence. Experimental results show that the incipience of the sensing-line blockage can be detected by this method.

Sinewave fit is a fundamental task in many test and measurement systems. The characterizations of analog–digital converters and digital oscilloscopes are two examples. In this paper, we present a high-performance (i.e., high-precision and high-speed) algorithm to estimate the four parameters of a sinewave from a sample data record. By the use of trigonometric identity, we propose a frequency estimator that turns the nonlinear estimation problem into a linear one. Thus, the difficulty of the traditional nonlinear least-squares sinewave fit method is attenuated. The total least-squares method is used to estimate four parameters of a sinewave in order to minimize the estimation errors in the sense of l2 norm. Simulation results exhibit that the proposed method gives superior performance over traditional ones and achieves excellent estimation of the true resolution of the simulated ideal ADC. This new algorithm is noniterative and gives swift and consistent results.

The development of a reliable, cost-effective instrument for on-line particle sizing remains a challenging area worldwide. This paper presents the design and evaluation of a new electrostatic sensor combined with a novel digital signal processing algorithm for the on-line continuous measurement of the mass median size of particles in a dilute-phase pneumatic suspension. Key design aspects of the constituent elements of the sensor, including a wire-mesh electrode, signal conditioning electronics and digital signal processing algorithm are described. Experimental results obtained using a typical range of particulate materials demonstrate that, under constant mass flow conditions, the processed output of the sensor is capable of indicating the mass median particle size with a relative error no greater than 15%. A critical appraisal of the technique, including advantages and limitations, is also included.