Using Wave Equation to Extract Digital Signal Features

Abstract

Voice signals are one of the most popular data types. They are used in various applications like security systems. In the current study a method based on wave equation was proposed, implemented and tested. This method was used for correct feature array generation. The feature array can be used as a key to identify the voice signal without any dependence on the voice signal type or size. Results indicated that the proposed method can produce a unique feature array for each voice signal. They also showed that the proposed method can be faster than other feature extraction methods.

A flexible method to create wave file features

“…Data histogram [20][21][22][23] is an array of elements, each of which points to the repetition of one value in the data set [24][25][26][27]. Calculating the wave file histogram is an initial task of the proposed later in this paper method of features extraction.…”

“…Here we have to select the data set, number of clusters, and the centroid of each cluster: Data set = 15,15,16,19,19,20,20,21,22,28,35,40,41,42,43,44,60,61,65 Clusters=2; C1=16, C2=22  Perform the following tasks while centroid changing:…”

Analysis of Objects Information to Create Unique Features for Color Images

“…Color images have a high resolution, thus they have a huge sizes which make the process of matching color images byte by byte a process that requires great effort and time, which force us to seek a method capable to generate a few unique values to represent any color image. Many texture methods are now used to create color image features [13], [14], many of these methods are based on local binary pattern (LBP) method such as modified LBP method, these methods provide high efficiency by requiring a small extraction time [15], [16], but these methods are sensitive to image rotation, any image rotation will generate new different features, broking the features stability condition [17]. Some of the methods used in extracting the image features depend on statistically treating the image matrix and calculating some statistical parameters such as the arithmetic mean and standard deviation, and using the values of these parameters as the image features.…”

Evaluation of speech signal features extraction methods

“…Microphones convert the fluctuating air pressure into electrical signals, voltages or currents, in which form we usually deal with speech signals in speech processing, speech signal is emerges from a speaker's mouth, nose and cheeks, is a one-dimensional function (air pressure) of time [1], [2], [20]. Microphones convert the fluctuating air pressure into electrical signals, voltages or currents, in which form we usually deal with speech signals in speech processing [17], [18], [19]. Human speech is an analogue signal which can be converted to digital signal by applying sampling and quantization as shown in figure 1.…”

“…The speech file histogram can be calculated based on local binary pattern (LBP) operator calculation [24], [25], and here we introduce the following method as shown in table 2 to calculate LBP histogram for each speech file. To reduce the number of values used to represent the speech signal file we have to seek a method to extract a set of features values [17], [18], [19,[20], which must be unique and small and easily used to identify the speech file.…”