Claims:1. A method for filtering a plurality of head movement noise of a user, the method comprising a processor implemented steps of:
acquiring one or more electrooculography (EOG) signals of a user;
filtering, using a first filter, the one or more acquired electrooculography (EOG) signals to obtain a first set of filtered electrooculography (EOG) signals;
smoothening, using a second filter, the first set of filtered electrooculography (EOG) signals to obtain one or more smoothened electrooculography (EOG) signals;
removing, one or more redundant patterns and one or more direct current (DC) drifts from the one or more smoothened electrooculography (EOG) signals to obtain a second set of filtered electrooculography (EOG) signals; and
applying, a discrete wavelet transform, on the second set of filtered electrooculography (EOG) signals to filter a plurality of head movement noise from the second set of filtered electrooculography (EOG) signals of the user.

2. The method of claim 1, wherein the step of applying, the discrete wavelet transform, on the second set of filtered electrooculography (EOG) signals comprises:
applying, on the second set of filtered electrooculography (EOG) signals, a mother wavelet transform and performing contracting, dilating, and shifting operations of the mother wavelet transform upon the second set of filtered electrooculography (EOG) signals to obtain a set of wavelets;
decomposing, at one or more decomposition levels, the set of wavelets to filter a plurality of head movement noise from the second set of filtered electrooculography (EOG) signals.

3. The processor implemented method of claim 1, wherein the step of removing one or more redundant patterns and one or more direct current (DC) drifts from the one or more smoothened electrooculography (EOG) signals comprises applying a nth order polynomial fitting on one or more vertical and horizontal channels of the one or more smoothened electrooculography (EOG) signals to obtain a best fitted polynomial, and subtracting the best fitted polynomial from the one or more smoothened electrooculography (EOG) signals to identify and remove the one or more redundant patterns and one or more direct current (DC) drifts.
4. A system for filtering a plurality of head movement noise of a user, the system comprising:
a memory storing instructions;
one or more communication interfaces; and
one or more hardware processors coupled to the memory via the one or more communication interfaces, wherein the one or more hardware processors are configured by the instructions to:
acquire one or more electrooculography (EOG) signals of a user;
filter using a first filter the one or more acquired electrooculography (EOG) signals to obtain a first set of filtered EOG signals;
smoothen using a second filter the first set of filtered electrooculography (EOG) signals to obtain one or more smoothened electrooculography EOG signals;
remove one or more redundant patterns and one or more direct current (DC) drifts from the one or more smoothened electrooculography (EOG) signals to obtain a second set of filtered electrooculography (EOG) signals; and
apply a discrete wavelet transform on the second set of filtered electrooculography (EOG) signals to filter a plurality of head movement noise from the second set of filtered electrooculography (EOG) signals of the user.

5. The system of claim 1, wherein the step of applying, the discrete wavelet transform, on the second set of filtered electrooculography (EOG) signals comprises:
applying, on the second set of filtered electrooculography (EOG) signals, a mother wavelet transform and perform contracting, dilating, and shifting operations of the mother wavelet transform upon the second set of filtered electrooculography (EOG) signals to obtain a set of wavelets;
decomposing, at one or more decomposition levels, the set of wavelets to filter plurality of head movement noise from the second set of filtered electrooculography (EOG) signals.

6. The system of claim 1, wherein the one or more redundant patterns and the one or more direct current (DC) drifts are removed from the one or more smoothened electrooculography (EOG) signals by applying a nth order polynomial fitting on one or more vertical and horizontal channels of the one or more smoothened electrooculography (EOG) signals to obtain a best fitted polynomial, and subtracting the best fitted polynomial from the one or more smoothened electrooculography (EOG) signals to identify and remove the one or more redundant patterns and one or more direct current (DC) drifts.
, Description:FORM 2

THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENT RULES, 2003

COMPLETE SPECIFICATION
(See Section 10 and Rule 13)

Title of invention:
SYSTEMS AND METHODS FOR WAVELET BASED HEAD MOVEMENT ARTIFACT REMOVAL FROM ELECTROOCULOGRAPHY (EOG) SIGNALS

Applicant:
Tata Consultancy Services Limited
A company Incorporated in India under the Companies Act, 1956
Having address:
Nirmal Building, 9th Floor,
Nariman Point, Mumbai 400021,
Maharashtra, India

The following specification particularly describes the invention and the manner in which it is to be performed.

TECHNICAL FIELD
[0001] The present application generally relates to the field of head movement noise removal from electrooculography (EOG) signals. More particularly, the present application relates to systems and methods for wavelet based head movement artifact removal from electrooculography (EOG) signals.

BACKGROUND
[0002] An electrooculogram is the electric potential measured around the eyes, which is generated by the corneo-retinal standing potential between the front and back of the eye. Pairs of electrodes are generally attached either to the left and right of the eyes (horizontal electrooculography component) or above and below the eye (vertical electrooculography component) to measure the eye movements. The horizontal and vertical electrooculography (EOG) components are then obtained by subtracting the signal obtained at one electrode from the signal at the other electrode. Electrooculography (EOG) signals acquire different types of eye movements, which can be employed for human-machine interfaces (HMI) and also for diagnostic purposes. Electrooculography (EOG) signals tend to be contaminated with noise due to unconstrained head movements. This head movement noise or artifact degrades the signal quality as well as increases the misclassification rate of eye movement detection. General filtering and preprocessing techniques are unable to remove this noise. Many traditional systems and methods have previously focused on the signal clarity but none of them have specifically focused on removing the head movement artifacts from the electrooculography signals.
[0003] Researchers generally carry out experiments in controlled lab environments, under constrained conditions so as to minimize any sort of contamination of the electrooculography (EOG) signals. Various factors may affect the electrooculography (EOG) signals quality which include power line noise, facial electromyography (EMG), loose electrode contact, and also head movement artifacts. Most of these artifacts can be removed by simple band pass, median, and/or moving average filtering. However the artifacts due to the head movements, in absence of chin rest or constraints of not moving the head, poses to be more problematic as it is in the same frequency range of electrooculography (EOG) signals and also morphologically close to electrooculography (EOG). Researchers have worked on removal of power line, blinks, and facial electromyography (EMG) noise from the electrooculography (EOG). To the best of authors’ knowledge none of the existing works have concentrated on presence of head-movement noise in electrooculography (EOG) signals and in turn removal of the same.
[0004] The artifacts due to the head movements, in absence of chin rest or constraints of not moving the head, poses to be problematic as it is in the same frequency range of electrooculography (EOG) signals and also morphologically close to electrooculography (EOG). Further, the electrooculography (EOG) signals contaminated with head movement noise tend to misclassification rate of eye movement recognition. Hence, there is a need for a technology that reduces the effect of head movement artifacts and thus to improve the classification accuracy of eye movement detection from electrooculography (EOG) signals.

SUMMARY
[0005] Before the present systems and methods, are described, it is to be understood that this application is not limited to the particular systems, and methodologies described, as there can be multiple possible embodiments which are not expressly illustrated in the present disclosures. It is also to be understood that the terminology used in the description is for the purpose of describing the particular versions or embodiments only, and is not intended to limit the scope of the present application. This summary is provided to introduce concepts related to wavelet based head movement artifact removal from electrooculography (EOG) signals and the concepts are further described below in the detailed description. This summary is not intended to identify essential features of the disclosure nor is it intended for use in determining or limiting the scope of the disclosure.
[0006] In an embodiment of the present disclosure, there is provided a method for wavelet based head movement artifact removal from electrooculography (EOG) signals, the method comprising: acquiring one or more electrooculography (EOG) signals of a user; filtering using a first filter, the one or more acquired electrooculography (EOG) signals to obtain a first set of filtered EOG signals; smoothening using a second filter, the first set of filtered electrooculography (EOG) signals to obtain one or more smoothened electrooculography (EOG) signals; removing one or more redundant patterns and one or more direct current (DC) drifts from the one or more smoothened electrooculography (EOG) signals to obtain a second set of filtered electrooculography (EOG) signals; applying a discrete wavelet transform on the second set of filtered electrooculography (EOG) signals to filter a plurality of head movement noise from the second set of filtered electrooculography (EOG) signals of the user by applying on the second set of filtered electrooculography (EOG) signals, a mother wavelet transform and performing contracting, dilating, and shifting operations of the mother wavelet transform upon the second set of filtered electrooculography (EOG) signals to obtain a set of wavelets and decomposing, at one or more decomposition levels, the set of wavelets to filter a plurality of head movement noise from the second set of filtered electrooculography (EOG) signals; and removing one or more redundant patterns and one or more direct current (DC) drifts from the one or more smoothened electrooculography (EOG) signals by applying a nth order polynomial fitting on one or more vertical and horizontal channels of the one or more smoothened electrooculography (EOG) signals to obtain a best fitted polynomial, and subtracting the best fitted polynomial from the one or more smoothened electrooculography (EOG) signals to identify and remove the one or more redundant patterns and one or more direct current (DC) drifts.
[0007] In an embodiment of the present disclosure, there is provided a system for wavelet based head movement artifact removal from electrooculography (EOG) signals, the system comprising: one or more processors; one or more data storage devices operatively coupled to the one or more processors and configured to store instructions configured for execution by the one or more processors to: acquire one or more electrooculography (EOG) signals of a user; filter using a first filter the one or more acquired electrooculography (EOG) signals to obtain a first set of filtered electrooculography (EOG) signals; smoothen using a second filter the first set of filtered electrooculography (EOG) signals to obtain one or more smoothened electrooculography (EOG) signals; remove one or more redundant patterns and one or more direct current (DC) drifts from the one or more smoothened electrooculography (EOG) signals to obtain a second set of filtered electrooculography (EOG) signals; apply a discrete wavelet transform on the second set of filtered electrooculography (EOG) signals to filter a plurality of head movement noise from the second set of filtered electrooculography (EOG) signals of the user; apply the discrete wavelet transform on the second set of filtered electrooculography (EOG) signals by applying on the second set of filtered EOG signals, a mother wavelet transform and perform contracting, dilating, and shifting operations of the mother wavelet transform upon the second set of filtered electrooculography (EOG) signals to obtain a set of wavelets and decompose at one or more decomposition levels, the set of wavelets to filter a plurality of head movement noise from the second set of filtered electrooculography (EOG) signals; and remove one or more redundant patterns and one or more direct current (DC) drifts from the one or more smoothened electrooculography (EOG) signals by applying a nth order polynomial fitting on one or more vertical and horizontal channels of the one or more smoothened electrooculography (EOG) signals to obtain a best fitted polynomial, and subtracting the best fitted polynomial from the one or more smoothened electrooculography (EOG) signals to identify and remove the one or more redundant patterns and one or more direct current (DC) drifts.

BRIEF DESCRIPTION OF THE DRAWINGS
[0008] The embodiments herein will be better understood from the following detailed description with reference to the drawings, in which:
[0009] Fig. 1 illustrates a block diagram of a system for wavelet based head movement artifact removal from electrooculography (EOG) signals;
[0010] Fig. 2 is a flowchart illustrating the steps involved for wavelet based head movement artifact removal from electrooculography (EOG) signals.
[0011] Fig. 3(a) illustrates visual representation of various eye-ball movements during electrooculography (EOG) acquisition;
[0012] Fig. 3(b) to 4(m) shows the graphical representation of the acquired and preprocessed electrooculography (EOG) signals during six types of eye movements; and
[0013] Fig. 4(a) to 4(e) shows the graphical representation of the set of electrooculography (EOG) signals obtained at each stage;

DETAILED DESCRIPTION OF THE EMBODIMENTS
[0014] The embodiments herein and the various features and advantageous details thereof are explained more fully with reference to the non-limiting embodiments that are illustrated in the accompanying drawings and detailed in the following description. The examples used herein are intended merely to facilitate an understanding of ways in which the embodiments herein may be practiced and to further enable those of skill in the art to practice the embodiments herein. Accordingly, the examples should not be construed as limiting the scope of the embodiments herein.
[0015] Referring now to the drawings, and more particularly to FIG. 1 through FIG 4, where similar reference characters denote corresponding features consistently throughout the figures, there are shown preferred embodiments and these embodiments are described in the context of the following exemplary system and/or method.
[0016] According to an embodiment of the disclosure, a block diagram of the system 100 is shown in Fig. 1. The system 100 includes one or more processors 104, communication interface device(s) or input/output (I/O) interface(s) 106, and one or more data storage devices or memory 102 operatively coupled to the one or more processors 104. The one or more processors 104 that are hardware processors can be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuitries, and/or any devices that manipulate signals based on operational instructions. Among other capabilities, the processor(s) is configured to fetch and execute computer-readable instructions stored in the memory. In an embodiment, the system 100 can be implemented in a variety of computing systems, such as laptop computers, notebooks, hand-held devices, workstations, mainframe computers, servers, a network cloud and the like.
[0017] The I/O interface device(s) 106 can include a variety of software and hardware interfaces, for example, a web interface, a graphical user interface, and the like and can facilitate multiple communications within a wide variety of networks N/W and protocol types, including wired networks, for example, LAN, cable, etc., and wireless networks, such as WLAN, cellular, or satellite. In an embodiment, the I/O interface device(s) can include one or more ports for connecting a number of devices to one another or to another server.
[0018] The memory 102 may include any computer-readable medium known in the art including, for example, volatile memory, such as static random access memory (SRAM) and dynamic random access memory (DRAM), and/or non-volatile memory, such as read only memory (ROM), erasable programmable ROM, flash memories, hard disks, optical disks, and magnetic tapes.
[0019] FIG 2 represents schematic flow illustrating a system and method for wavelet based head movement artifact removal from electrooculography (EOG) signals according to an embodiment of the present disclosure. At step 201, one or more Electrooculography (EOG) signals of a user are acquired. Initially, at step 202, the one or more acquired electrooculography (EOG) signals to obtain a first set of filtered electrooculography (EOG) signals are filtered using a first filter (not shown in FIGS. 1-2). At step 203, the first set of filtered electrooculography (EOG) signals to obtain one or more smoothened electrooculography (EOG) signals are smoothened using a second filter (not shown in FIGS. 1-2). At step 204, removing one or more redundant patterns and one or more DC drifts from the one or more smoothened electrooculography (EOG) signals to obtain a second set of filtered electrooculography (EOG) signals may be performed according to an embodiment of the present disclosure. At step 205, a discrete wavelet transform on the second set of filtered electrooculography (EOG) signals is applied to filter the head movement artifacts from the second set of filtered electrooculography (EOG) signals of the user.
[0020] According to an embodiment, the electrooculography (EOG) acquisition system comprises a two channel acquisition system, one each for vertical and horizontal eye movement signals and if further universal serial bus (USB) powered, with a provision of signal isolation achieved by direct current DC/DC converter, which isolates the circuit from USB 5V power and generates a ±12V, which further powers the rest of the circuit. The amplitude and frequency range of electrooculography (EOG) signals are 5-30µV and 0.01-20Hz respectively. To avoid signal saturation due to noise, the developed circuit has a total gain of 2400 given in three stages. The circuit further comprises an instrumentation preamplifier with a gain G1=100 where the output of said preamplifier acts an input to passive high pass filter with low cut off frequency of 0.1Hz, that reduces the direct current (DC) drifts. The high pass filter (HPF) is followed by an active low pass filter of high cut off of 40Hz and again G2=2.4. The circuit further comprises of an amplifier with a gain G3=10. The electrooculography (EOG) signals are transmitted to PC through 16bit analogue to digital converter (ADC), National Instruments Universal Serial Bus (NI USB) 6216. The two channel circuit has a current consumption of 9mA. The sampling rate is 100Hz. Ag/AgCl electrodes are utilized. The signals are further processed in matrix laboratory (MATLAB) environment, a software platform. The raw electrooculography (EOG) signal for right and left eye movement are depicted in FIG 2(b), electrooculography vertical channel (EOG_V) in cyan and electrooculography horizontal channel (EOG_H) in green.
[0021] FIG 3 (a) illustrates visual representation of various eye-ball movements during electrooculography (EOG) acquisition. The electrooculography (EOG) signals acquired comprises of plurality of movements (six movements) types voluntary blinks, single (SB) and double (DB), and up (U), down (D), left (L) and right (R) directions with and without head movement. According to an embodiment of the disclosure, the electrooculography (EOG) system may be utilized to acquire data from few users comprising 4 males and 2 females (27±5years). The users are first seated at a distance of 120cm in front of a 17” rectangular computer screen (height: 10inch, width: 13inch, approximately), and are presented to visual cues. Referring to FIG 3(a) again, the acquisition of electrooculography (EOG) signals with and without head movement noise may be then be performed according to an embodiment, by marking the origin of the user by fixing a cross in the middle of the screen for few seconds; moving the ball of 60 pixel size in the middle of the screen; then the ball either moved up/down/left/right and the user instructed to follow the ball with eyes; and the sequence of ball movement in any direction followed the pattern: center (5 seconds)->left/right/up/down(1sec)-> center (5seconds). For each direction there were 20 trials per session, and there were total two sessions, thus total 40 trials for each movement type per subject. In addition to this the instruction “BLINK” or “BLINK TWICE” (in font size 97) also appeared on screen, each 20 times in a session, where the subjects needed to blink voluntarily, once or twice, respectively.
[0022] Fig. 3(b) to 3(m) shows the graphical representation of the acquired and preprocessed electrooculography (EOG) signals during six types of eye movements. In FIG. 3(b) to 3(m), the dotted line denotes EOGH (horizontal channel EOG) and the solid line denotes EOGV (vertical channel EOG). FIG. 3(b) shows single eye blink and may be observed as a positive spike more dominant in the vertical EOG channel. This does not contain head movement noise. FIG. 3(c) shows single eye blink movement and may be observed as a single positive spike in vertical channel (towards the end of the graph). The deflections at the beginning, in the vertical channel denotes the head movement noise. FIG 3(d) shows double eye blinks (consequent two eye blinks) and may be observed as two positive spikes more dominant in the vertical EOG channel. This does not contain head movement noise. FIG 3(e) shows double eye blink movement and may observed to be as two positive spikes in vertical channel. The other spikes in the vertical channel denotes the head movement noise. FIG 3(f) shows left eye movement and may be observed as a positive pulse like deflection in the horizontal EOG channel, this is devoid of any head movement noise. FIG 3(g) shows left eye movement, as a positive pulse like deflection in the horizontal EOG channel. The big deflections in the vertical channel (towards the end of the graph) denotes the head movement noise. FIG 3(h) shows right eye movement as a negative pulse deflection seen in the horizontal channel. This does not contain head movement noise. FIG 3(i) shows right eye movement as a negative pulse deflection seen in the horizontal channel. The big deflections in the vertical channel (towards the end of the graph) denotes the head movement noise. FIG 3(j) shows the up eye movement, as positive pulse like deflection in the vertical EOG channel. It does not contain any head movement noise. FIG 3(k) shows the up eye movement, as positive pulse like deflection in the vertical EOG channel (at the beginning of the graph). The other big deflections in the vertical channel denote the head movement noise. FIG 3(l) shows the down eye movement, as a negative pulse like deflection in the vertical channel. It does not contain any head movement noise. FIG 3(m) shows down eye movement, as a very small negative pulse like deflection in the vertical channel, towards the beginning of the graph. All other big deflections in the vertical channel towards the middle and end shows head movement noise.
[0023] FIG. 4(a) to 4(e) provides graphical representation of the electrooculography (EOG) signals obtained at each stage of the disclosure (from the acquisition till head movement noise removal) may be observed according to an embodiment of the disclosure. Initially, the electrooculography (EOG) signals of a user are acquired using the electrooculography (EOG) acquisition system. These signals are the raw signals acquired with head movement artifacts. Referring to FIG. 4(a) the acquired electrooculography (EOG) signal of horizontal channel, consisting of left and right eye movements and also head movement noise may be observed. After the acquisition of the electrooculography (EOG) signals, these are filtered using the first filter to obtain first set of electrooculography (EOG) signals which still contain head movement artifacts. The first filter and the second filter are executed by the system 100 and may be stored in the memory 102. According to an embodiment the filtration to obtain first set of electrooculography (EOG) signals may be performed by applying 4th order FIR bandpass filter (referred herein as a first filter) with Hamming window. Referring to FIG. 4(b) the filtered electrooculography (EOG) signals obtained after applying the first filter (4th order FIR bandpass filter) may be observed. After obtaining first set of filtered electrooculography (EOG) signals, these are further smoothened using a second filter (1-dimensional, 4th order median filter) to obtain smoothened electrooculography (EOG) signals. According to an embodiment this may be performed by applying 1-dimensional, 4th order median filter (referred herein as the second filter), which smoothens the signal. Referring to FIG. 4(c), the smoothened electrooculography (EOG) signals may be observed. From the smoothened electrooculography (EOG) signals, redundant patterns and DC drifts are removed to obtain further filtered electrooculography (EOG) signals. According to an embodiment of the present disclosure this may be performed by applying a nth order (eg 6th order as discussed below) polynomial fitting on one or more vertical and horizontal channels of the one or more smoothened electrooculography (EOG) signals to obtain a best fitted polynomial, and subtracting the best fitted polynomial from the one or more smoothened electrooculography (EOG) signals to identify and remove the one or more redundant patterns and one or more DC drifts. Referring to FIG. 4(d), the electrooculography (EOG) signals after removing redundant patterns and DC drafts may be observed. However these signals still contain head movement artifacts. Finally, the electrooculography (EOG) signals are obtained by applying discrete wavelet transform to the electrooculography (EOG) signals obtained after removing redundant patterns and DC drifts. These electrooculography (EOG) signals do not contain any head movement artifacts. Referring to FIG. 4(e), the final electrooculography (EOG) signals without any head movement artifacts may be observed.
[0024] According to an embodiment of the disclosure, the removal of a plurality of head movement noise from the acquired electrooculography (EOG) signals may now be considered in detail. The above acquired electrooculography (EOG) signals are filtered, using a first filter to obtain a first set of filtered electrooculography (EOG) signals. In an embodiment, the filtration of the acquired electrooculography (EOG) signals may be performed using 4th order FIR bandpass filter with Hamming window according to the equation (1):
, and
=0.54-0.46 equation (1)
Where, is the Hamming window function of finite duration, is the practical FIR filter, desired IIR filter prototype and is the filter order. The lower and upper cut-off frequencies are set as 0.5 and 20Hz respectively.
[0025] According to an embodiment of the disclosure, the first set of filtered electrooculography (EOG) signals, also referred to as bandpass filtered signal are further smoothened, by using a second filter to obtain one or more smoothened electrooculography (EOG) signals. In an embodiment this may be performed by applying 1-dimensional, 4th order median filter, which smoothens the signal and at the same time preserves distinctive edges. However, the smoothened electrooculography (EOG) signals still contain some direct current (DC) drifts and redundant patterns as these non-linear patterns and direct current (DC) drifts are present throughout in the acquired electrooculography (EOG) signals initially and hence they are required to be removed from obtained smoothened electrooculography (EOG) signals to obtain a second set of further filtered electrooculography (EOG) signals to avoid any glitches, inaccuracy in further processing and analysis. The redundant or non-linear patterns and direct current (DC) drifts are removed from said smoothened electrooculography (EOG) signals by applying a 6th order polynomial fitting separately to the median filtered vertical (EOG_V) and horizontal (EOG_H) channels of electrooculography (EOG) and by further subtracting the best fitted polynomial from the one or more smoothened electrooculography (EOG) signals to identify and remove the one or more redundant patterns and one or more direct current (DC) drifts. This further provides a second set of further filtered electrooculography (EOG) signals. Referring to FIG. 4(e) again, the final electrooculography (EOG) signals obtained without head movement artifacts may be observed.
[0026] According to an embodiment of the disclosure, the removal of a plurality of head movement noise signals further comprises applying a discrete wavelet transform on the second set of filtered electrooculography (EOG) signals to filter a plurality of head movement noise from the second set of filtered electrooculography (EOG) signals of the user. However, prior to performing discrete wavelet analysis, an eye movement epochs are extracted from the second set of filtered electrooculography (EOG) signals, where each epoch was of 4 seconds window. During real time processing (i.e. online classification), then instead of epoch extraction, signals may be buffered for each 4 seconds. Thus it helps in signal buffering instead of repeated epoch extractions. To avoid demerit of fixed window lengths, the present disclosure applies a discrete wavelet transform as it can discriminate between time and frequency domain characteristics. The step of applying the discrete wavelet transform on the second set of filtered electrooculography (EOG) signals further comprises: applying, on the second set of filtered electrooculography (EOG) signals, a mother wavelet transform (also referred herein as single archetype wavelet transform) and performing contracting, dilating, and shifting operations of a mother wavelet (a single archetype wavelet) of the mother wavelet transform upon the second set of filtered electrooculography (EOG) signals to obtain a set of wavelets; and decomposing, at one or more decomposition levels, the set of wavelets to filter a plurality of head movement noise from the second set of filtered electrooculography (EOG) signals. The single archetype wavelet, referred to as the mother wavelet is subjected to contraction, dilation, and shifting operations to obtain wavelets. The obtained wavelets are the origin functions that are segregated with respect to time and frequency and are further used to decompose, at one or more decomposition levels, the set of wavelets to filter a plurality of head movement noise from the second set of filtered electrooculography (EOG) signals. This approach forms the basis of wavelet transformation. Mother wavelet may be represented by eq. (2) as:
equation (2)
where is the scaling factor and is the shifting parameter and is the wavelet space.
[0027] As the maximum power of the acquired electrooculography (EOG) signals are contained below 15Hz, the present disclosure perform the decomposition of the set of wavelets to filter head movement noise from the second set of filtered electrooculography (EOG signals) till level 4. The level 4 decomposed wavelets remove the head movement noise and also further retain the signal morphology related to various eye movements. Further, the present disclosure implements biorthogonal ‘bior2.8’ mother wavelet or single archetype wavelet for decomposing the set of wavelets to filter head movement noise from the second set of filtered electrooculography (EOG) signals as it resembles the eye movements very closely. The biorthogonal wavelet further also provide more degrees of freedom and orthogonal counterparts. The present disclosure applies the discrete wavelet transform for both head movement noise filtering and feature extraction.
[0028] The present disclosure also facilitates reduction in signal reconstruction time and computational load as it performs the wavelet transformation based decomposition only and utilizes the obtained decomposed or approximated coefficients for further processing. The present disclosure exploits there types of features, feature set 1 (FS1): the obtained decomposition / approximation coefficients serve as feature set to the classifier, feature set 2 (FS2): sometime-domain parameters and statistical parameters (viz. area under the curve, peak to peak amplitude, maximum and minimum value in a particular epoch window, Hjorth parameters, standard deviation, mean, Skewness, Kurtosis, Shannon’s entropy) are extracted from the level 4 approximation coefficients, and iii) feature set 3 (FS3): same time domain and statistical parameters as extracted from preprocessed electrooculography (EOG), prior to wavelet decomposition. These are separately classified. Referring to FIG 2(b) electrooculography (EOG_V) denotes the preprocessed and filtered electrooculography EOG.
[0029] According to an embodiment of the present disclosure, classification of eye movements from extracted feature sets may be considered. Classification of plurality of eye movements (eg. six eye movements) from the extracted feature sets FS1, FS2 and FS3 are carried out by multiclass k-nearest neighbor (kNN) [21], with k=5 and Euclidean distance as the distance metric Classification results, in the form of confusion matrices (CM), averaged over all subjects for features sets FS1, FS2 and FS3 for 40 trials for each movement type are depicted in Table 1, 2 and 3 respectively. The runtime for classification using feature set FS1, FS2, and FS3 is 0.11 sec, 4.72 sec and 2.55 sec respectively. Referring to tables 1, 2 and 3 below, the system performed best with FS1 followed by FS2 and FS3, moreover FS1 took the least time. Thus the proposed wavelet transform (WT) based denoising as well as feature extraction improves the performance of eye movement recognition system.
[0030] Referring to table 1, 2 and 3 below, different types of eye movement are denoted by ‘R’ (tight eye movement), ‘L’ (left eye movement), ‘U’ (up eye movement), ‘D’ (down eye movement), ‘SB’ (single blink) and ‘DB’ (double blink). The table 1 confusion matrices have been obtained while classifying with feature set 1 (FS1). Similarly, table 2 and table 3 confusion matrices have been obtained while classifying with feature set 2 (FS2) and feature set 3 (FS3) respectively. The figures in bold denote the maximum classification accuracy of particular eye movement class among all three feature sets (FS1, FS2 and FS3).
Table 1: Confusion matrice (CM) for feature set 1 (FS1)
CLASSES PREDICTED
R L U D SB DB
TRUE R 30 0 0 8 2 0
L 0 26 1 12 0 1
U 2 2 30 5 0 1
D 1 0 1 37 0 1
SB 1 0 5 9 25 0
DB 0 0 9 2 10 19

Table 2: Confusion matrice (CM) for feature set 2 (FS2)
CLASSES PREDICTED
R L U D SB DB
TRUE R 28 0 1 8 0 3
L 9 13 3 13 1 1
U 3 1 22 13 0 1
D 0 3 2 34 0 1
SB 5 2 0 1 23 9
DB 3 2 1 0 3 31

Table 3: Confusion matrice (CM) for feature set 3 (FS3)
CLASSES PREDICTED
R L U D SB DB
TRUE R 18 11 1 4 0 6
L 8 27 1 1 1 2
U 3 0 18 15 1 3
D 2 0 3 30 2 3
SB 0 0 0 2 33 5
DB 2 1 0 2 4 31
[0031] According to an embodiment of the present disclosure, the comparison of the present disclosure with related traditional systems and methods may be considered. The traditional systems and methods fail to consider the head movement noise removal or filtering specifically for acquiring different kinds of eye movements using electrooculography (EOG). Further, the present disclosure has applied discrete wavelet transform for both head movement noise filtering and feature extraction which has significantly improved performance of acquiring different kinds of eye movements using electrooculography (EOG). Referring to table 4 below, Traditional systems and methods 1, Traditional systems and methods 3 and Traditional systems and methods 5 when applied on the present data set, without and with head movement noise have been presented. In Traditional systems and methods 2, Traditional systems and methods 4 and Traditional systems and methods 6, prior to applying Traditional systems and methods 1, Traditional systems and methods 3 and Traditional systems and methods 5 respectively, present disclosure wavelet (WT) based denoising has been implemented, which has improved the classification accuracies (CA) in each of the cases, and the standard deviations (SD) are given in parenthesis have decreased. The bold figures in Table 4 below denote the best classification accuracies. Moreover, unlike other wavelet transform approaches, the present disclosure have utilized the approximation coefficients only. This reduces the time and computational complexity as discussed previously.
Table 4 – Performance of the present disclosure compared with traditional systems and methods
Classification accuracy (CA) in % standard deviations (SD)
Traditional systems and methods Without head movement Head movement
Traditional systems and methods 1 86 (2.2) 81.6 (2.7)
Traditional systems and methods 2: WT (present disclosure wavelet) + Traditional systems and methods 1 98.3 (0.9) 93.3 (1.8)
Traditional systems and methods 3 92.5 (8.6) 56.2 (4.6)
Traditional systems and methods 4: WT (present disclosure wavelet) + Traditional systems and methods 3 95 (7.07) 72.5 (3.3)
Traditional systems and methods 5 77.6 (4.7) 72.3 (4.8)
Traditional systems and methods 6: WT (present disclosure wavelet) + Traditional systems and methods 5 75 (8.3) 80.5 (4.8)
[0032] The present disclosure considers plurality of eye movements (eg. six types) mentioned above acquired with and without head movement noise. The electrooculography (EOG) signals contaminated with head movement noise tend to misclassification rate of eye movement recognition. The decomposed electrooculography (EOG) signals are extracted by applying discrete wavelet transform to filter head movement artifacts as well as to increase the accuracy of eye movement classification. The present disclosure can be implemented in real time systems as well. The present disclosure when compared with related traditional systems and methods, increase the accuracy of existing works as well as shown in the comparison above.
[0033] It is, however to be understood that the scope of the protection is extended to such a program and in addition to a computer-readable means having a message therein; such computer-readable storage means contain program-code means for implementation of one or more steps of the method, when the program runs on a server or mobile device or any suitable programmable device. The hardware device can be any kind of device which can be programmed including e.g. any kind of computer like a server or a personal computer, or the like, or any combination thereof. The device may also include means which could be e.g. hardware means like e.g. an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a combination of hardware and software means, e.g. an ASIC and an FPGA, or at least one microprocessor and at least one memory with software modules located therein. Thus, the means can include both hardware means and software means. The method embodiments described herein could be implemented in hardware and software. The device may also include software means. Alternatively, the embodiments may be implemented on different hardware devices, e.g. using a plurality of CPUs.
[0034] The embodiments herein can comprise hardware and software elements. The embodiments that are implemented in software include but are not limited to, firmware, resident software, microcode, etc. The functions performed by various modules described herein may be implemented in other modules or combinations of other modules. For the purposes of this description, a computer-usable or computer readable medium can be any apparatus that can comprise, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
[0035] The medium can be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system (or apparatus or device) or a propagation medium. Examples of a computer-readable medium include a semiconductor or solid state memory, magnetic tape, a removable computer diskette, a random access memory (RAM), a read-only memory (ROM), a rigid magnetic disk and an optical disk. Current examples of optical disks include compact disk-read only memory (CD-ROM), compact disk-read/write (CD-R/W) and DVD.
[0036] A data processing system suitable for storing and/or executing program code will include at least one processor coupled directly or indirectly to memory elements through a system bus. The memory elements can include local memory employed during actual execution of the program code, bulk storage, and cache memories which provide temporary storage of at least some program code in order to reduce the number of times code must be retrieved from bulk storage during execution.
[0037] Input/output (I/O) devices (including but not limited to keyboards, displays, pointing devices, etc.) can be coupled to the system either directly or through intervening I/O controllers. Network adapters may also be coupled to the system to enable the data processing system to become coupled to other data processing systems or remote printers or storage devices through intervening private or public networks. Modems, cable modem and Ethernet cards are just a few of the currently available types of network adapters.
[0038] A representative hardware environment for practicing the embodiments may include a hardware configuration of an information handling/computer system in accordance with the embodiments herein. The system herein comprises at least one processor or central processing unit (CPU). The CPUs are interconnected via system bus to various devices such as a random access memory (RAM), read-only memory (ROM), and an input/output (I/O) adapter. The I/O adapter can connect to peripheral devices, such as disk units and tape drives, or other program storage devices that are readable by the system. The system can read the inventive instructions on the program storage devices and follow these instructions to execute the methodology of the embodiments herein.
[0039] The system further includes a user interface adapter that connects a keyboard, mouse, speaker, microphone, and/or other user interface devices such as a touch screen device (not shown) to the bus to gather user input. Additionally, a communication adapter connects the bus to a data processing network, and a display adapter connects the bus to a display device which may be embodied as an output device such as a monitor, printer, or transmitter, for example. The preceding description has been presented with reference to various embodiments. Persons having ordinary skill in the art and technology to which this application pertains will appreciate that alterations and changes in the described structures and methods of operation can be practiced without meaningfully departing from the principle, spirit and scope.