Arindam Jati

Object-shape recognition from tactile images using a feed-forward neural network

The sense of touch is an extremely important sensory system in the human body which helps to understand object shape, texture, hardness in the world around us. Incorporating artificial haptic sensory systems in rehabilitative aids and in various other human computer interfaces is a thrust area of research presently. This paper presents a novel approach of shape recognition and classification from the tactile pressure images by touching the surface of various real life objects. Here four objects (viz. a planar surface, object with one edge, a cuboid i.e. object with two edges and a cylindrical object) are used for shape recognition. The obtained tactile pressure images of the object surfaces are subjected to segmentation, edge detection and a mapping procedure to finally reconstruct the particular object shapes. The reconstructed images are used as features. The processed tactile pressure images are classified with feed- forward neural network (FFNN) using extracted features. The classifier performance is tested with different signal-to-noise (SNR) ratios. Is is observed that classifier accuracy decreases with decrease in SNR, but at SNR value 6 i.e. when the noise power is one sixth of the signal power, the mean classification accuracy of the classifier is 88%. This shows the robustness of feed-forward neural network in the classification purpose. The performance of FFNN is compared with four classifiers (Linear Discriminant Analysis, Linear Support vector machine, Radial Basis Function SVM, k-Nearest Neighbor). FFNN performed best acquiring first rank with a average classification accuracy of 94.0%.

Automatic leukocyte nucleus segmentation by intuitionistic fuzzy divergence based thresholding

The paper proposes a robust approach to automatic segmentation of leukocytes nucleus from microscopic blood smear images under normal as well as noisy environment by employing a new exponential intuitionistic fuzzy divergence based thresholding technique. The algorithm minimizes the divergence between the actual image and the ideally thresholded image to search for the final threshold. A new divergence formula based on exponential intuitionistic fuzzy entropy has been proposed. Further, to increase its noise handling capacity, a neighborhood-based membership function for the image pixels has been designed. The proposed scheme has been applied on 110 normal and 54 leukemia (chronic myelogenous leukemia) affected blood samples. The nucleus segmentation results have been validated by three expert hematologists. The algorithm achieves an average segmentation accuracy of 98.52% in noise-free environment. It beats the competitor algorithms in terms of several other metrics. The proposed scheme with neighborhood based membership function outperforms the competitor algorithms in terms of segmentation accuracy under noisy environment. It achieves 93.90% and 94.93% accuracies for Speckle and Gaussian noises, respectively. The average area under the ROC curves comes out to be 0.9514 in noisy conditions, which proves the robustness of the proposed algorithm.

Object shape recognition from tactile images using regional descriptors

This paper presents a novel approach of shape recognition from the tactile images by touching the surface of various real life objects. Here four geometric shaped objects (viz. a planar surface, object with one edge, a cubical object i.e. object with two edges and a cylindrical object) are used for shape recognition. The high pressure regions denoting surface edges have been segmented out via multilevel thresholding. These high pressure regions hereby obtained were unique to different object classes. Some regional descriptors have been used to uniquely describe the high pressure regions. These regional descriptors have been employed as the features needed for the classification purpose. Linear Support Vector Machine (LSVM) classifier is used for object shape classification. In noise free environment the classifier gives an average accuracy of 92.6%. Some statistical tests have been performed to prove the efficacy of the classification process. The classifier performance is also tested in noisy environment with different signal-to-noise (SNR) ratios.

Object-shape classification and reconstruction from tactile images using image gradient

A human explores the world around him through his sense to touch. Touch sensation enables us to understand shape, texture and hardness of an object/surface necessary for efficient exploration. Incorporating artificial haptic sensory systems in rehabilitative aids and in various other human computer interfaces enhances the dexterity. This paper presents a novel approach of shape reconstruction and classification from the tactile images by touching the surface of various real life objects. Here four objects (viz. a planar surface, object with one edge, a cuboid i.e. object with two edges and a cylindrical object) have been used for the shape recognition purpose. A new gradient based feature extraction technique has been used for the classification purpose. The reconstruction algorithm also uses image gradients to differentiate between a surface having continuous curvature and a surface having sharp edge. Prewitt masks are used for determining the gradients. A comparison between the performances of different classifiers has been drawn to prove the efficacy of the shape classification algorithm.

A novel segmentation approach for noisy medical images using Intuitionistic fuzzy divergence with neighbourhood‐based membership function

Medical image segmentation demands higher segmentation accuracy especially when the images are affected by noise. This paper proposes a novel technique to segment medical images efficiently using an intuitionistic fuzzy divergence–based thresholding. A neighbourhood‐based membership function is defined here. The intuitionistic fuzzy divergence–based image thresholding technique using the neighbourhood‐based membership functions yield lesser degradation of segmentation performance in noisy environment. Its ability in handling noisy images has been validated. The algorithm is independent of any parameter selection. Moreover, it provides robustness to both additive and multiplicative noise. The proposed scheme has been applied on three types of medical image datasets in order to establish its novelty and generality. The performance of the proposed algorithm has been compared with other standard algorithms viz. Otsus method, fuzzy C‐means clustering, and fuzzy divergence–based thresholding with respect to (1) noise‐free images and (2) ground truth images labelled by experts/clinicians. Experiments show that the proposed methodology is effective, more accurate and efficient for segmenting noisy images.

Facial Action Point Based Emotion Recognition by Principal Component Analysis

This paper proposes an alternative approach to emotion recognition of a subject from selected 36 facial action points marked at specific locations on their faces. Facial expressions obtained by the subjects enacting them are recorded, and the corresponding changes in marked action points are measured. The measurements reveal that the action points have wider variations in facial expressions containing diverse emotions. Considering 10 instances for each facial expression, and carrying the same emotion, experimented over 10 subjects, we obtain a set of 100 distance matrices, representing the distance between any two selected action points. The 100 matrices for each individual emotion are averaged, and the first principal component, representing the most prominent features of the average distance matrix is evaluated. During the recognition phase, the first Principal component obtained from the distance matrix of an unknown facial expression is evaluated, and its Euclidean distance with the first Principal component of each emotion is determined. The unknown facial expression is classified into emotion class j, if the Euclidean distance between the obtained principal component and that of j-th emotion class is minimum. Classification of 120 facial images, containing equal number of samples for six emotion classes, reveals an average classification accuracy of 92.5%, the highest being in relax and disgust and the least in fear and anger.

A hybridisation of Improved Harmony Search and Bacterial Foraging for multi-robot motion planning

This paper provides a new approach to include the chemotactic behavior of Bacterial Foraging Algorithm (BFOA) in the existing Improved Harmony Search (IHS) algorithm. Extensive computer simulations with CEC-2005 benchmark functions reveal that the proposed algorithm outperforms the existing one with respect to accuracy in determining the optima. The proposed algorithm has successfully been implemented for multi-robot motion planning application. Performance has been studied using the proposed IHS-BFO algorithm and compared with existing IHS and Particle Swarm Optimization (PSO) algorithm.

OBJECT-SHAPE RECOGNITION BY TACTILE IMAGE ANALYSIS USING SUPPORT VECTOR MACHINE

The sense of touch is important to human to understand shape, texture, and hardness of the objects. An object under grip, i.e. object exploration by enclosure, provides a unique pressure distribution on the different regions of palm depending on its shape. This paper utilizes the above experience for recognition of object shapes by tactile image analysis. The high pressure regions (HPRs) are segmented and analyzed for object shape recognition rather than analyzing the entire image. Tactile images are acquired by capacitive tactile sensor while grasping a particular object. Geometrical features are extracted from the chain codes obtained by polygon approximation of the contours of the segmented HPRs. Two-level classification scheme using linear support vector machine (LSVM) is employed to classify the input feature vector in respective object shape classes with an average classification accuracy of 93.46% and computational time of 1.19 s for 12 different object shape classes. Our proposed two-level LSVM reduces the misclassification rates, thus efficiently recognizes various object shapes from the tactile images.

Negative emotion recognition from stimulated EEG signals

This paper proposes a scheme of emotion recognition from audio-visually stimulated EEG (electroencephalography) signals, the choice of signal being influenced by the fact that these signals are the direct unaltered outcome of ones brain activity and hence cannot be voluntarily suppressed. The EEG signals have been recorded using the NEUROWIN EEG amplifier of Nasan Medicals with a sampling rate of 250Hz from electrodes positioned at F3, F4, Fp1 and Fp2, since they lie over the frontal and pre-frontal lobe. The Raw EEG signals obtained need to be processed and classified into different emotional categories, using various features and intelligent classification algorithms. Features from these signals have been extracted using wavelet transform, statistical parameters and Hjorth parameter estimation, which are then classified using linear support vector machine (LSVM) and k-nearest neighbour (kNN). These extracted features are classified into the two different negative emotion classes of sad and disgust, with an average classification accuracy of the sad emotion being 78.04% and disgust being 76.31%. With our objective of development of emotionally challenged machines and devices that could become compatible with the emotional state of the user and nullify the effects of negative emotions on their work performance; the proposed scheme takes us a step closer to realisation of the same.

A Comparative Analysis of Emotion Recognition from Stimulated EEG Signals

This paper proposes a scheme to utilize the unaltered direct outcome of brain’s activity viz. EEG signals for emotion detection that is a prerequisite for the development of an emotionally intelligent system. The aim of this work is to classify the emotional states experimentally elicited in different subjects, by extracting their features for the alpha, beta, and theta frequency bands of the acquired EEG data using PSD, EMD, wavelet transforms, statistical parameters, and Hjorth parameters and then classifying the same using LSVM, LDA, and kNN as classifiers for the purpose of categorizing the elicited emotions into the emotional states of neutral, happy, sad, and disgust. The experimental results being a comparative analysis of the different classifier performances equip us with the best accurate means of emotion recognition from the EEG signals. For all the eight subjects, neutral emotional state is classified with an average classification accuracy of 81.65 %, highest among the other three emotions. The negative emotions including sad and disgust have better average classification accuracy of 76.20 and 74.96 % as opposed to the positive emotion i.e., happy emotional state, the average classification accuracy of which turns out to be 73.42 %.