International Journal of Image and Graphics

Novel high-speed memory optimized distributed arithmetic (DA)-based architecture is developed and modeled for 3D discrete wavelet transform (DWT). The memory requirement for the proposed architecture is designed to $9\times 9N+36$ pixel dynamic memory space and $52W$ ROM. The proposed 3D-DWT architecture implements 9/7 Daubechies wavelet filters, synthesizes 7127 bytes of memory for temporary storage and uses 758 adders, 36 multiplexers of 16:1 and 36 up counter to realize the 3D-DWT hardware. The 3D-DWT engine is implemented and tested in a Xilinx FPGA Vertex5 XC5VLX155T with high area and power efficiency. The maximum delay in the timing path is 2.676ns and the 3D-DWT works at maximum frequency of 381MHz clock.

In the domain of computer technology, image processing strategies have become a part of various applications. A few broadly used image segmentation methods have been characterized as seeded region growing (SRG), edge-based image segmentation, fuzzy $k$-means image segmentation, etc. SRG is a quick, strongly formed and impressive image segmentation algorithm. In this paper, we delve into different applications of SRG and their analysis. SRG delivers better results in analysis of magnetic resonance images, brain image, breast images, etc. On the other hand, it has some limitations as well. For example, the seed points have to be selected manually and this manual selection of seed points at the time of segmentation brings about wrong selection of regions. So, a review of some automatic seed selection methods with their advantages, disadvantages and applications in different fields has been presented.

Emotion recognition is becoming commercially popular due to the major role of analytics in various aspects of marketing and strategy management. Several papers have been proposed in emotion recognition. They are mainly classified in the past under 2D and 3D emotion recognition, out of which 2D emotion recognition has been more popular. Various aspects like facial posture, light intensity variations and sensor-independent recognition have been studied by different authors in the past. However, in reality, 3D emotion recognition has been found to be more efficient which has a broader area of use. In this paper, a 3D tracking plane with 2D feature points has enabled us to recognize emotions by statistical voting method from all planes having over threshold number of points in their respective contour area. The proposed technique’s results are comparable to existing methods in terms of time, space complexity and accuracy improvement.

Images are widely used in engineering. Unfortunately, ultrasound images are mainly degraded by an intrinsic noise called speckle. Therefore, de-speckling is a critical preprocessing step. Therefore, a robust despeckling method and accurate evaluation of images are suggested. We suggest three phases and a three-step denoising filter. In the first phase, the coefficients of variation are computed from the noisy image. The second phase is a three-step denoising filter. The first step is denoising of extreme levels of homogeneous regions, based on fuzzy homogeneous regions. The second step is a proposed adaptive bilateral filter (ABF). The ABF helps for better denoising based on the three regions which are edge, detail and homogeneous regions. The next step, a weight, is applied to the ABF. This step is for isolated noise denoising. Next, in the third phase, the output image is evaluated by the fuzzy logic approach. The proposed method is compared with other filters in the literature. The experimental outcomes show that the proposed method has better performance than the other filters. That proposed denoising algorithm is able to preserve image details and edges when compared with other denoising methods.

Efficient skin lesion segmentation algorithms are required for computer aided diagnosis of skin cancer. Several algorithms were proposed for skin lesion segmentation. The existing algorithms are short of achieving ideal performance. In this paper, a novel semi-automatic segmentation algorithm is proposed. The fare concept of the proposed is 8-directional search based on threshold for lesion pixel, starting from a user provided seed point. The proposed approach is tested on 200 images from PH2 and 900 images from ISBI 2016 datasets. In comparison to a chosen set of algorithms, the proposed approach gives high accuracy and specificity values. A significant advantage of the proposed method is the ability to deal with discontinuities in the lesion.

With the increasing demand for security in many fastest growing applications, biometric recognition is the most prominent authentication system. User authentication through speech and face recognition is the important biometric technique to enhance the security. This paper proposes a speech and facial feature-based multi-modal biometric recognition technique to improve the authentication of any system. Mel Frequency Cepstral Coefficients (MFCC) is extracted from audio as speech features. In visual recognition, this paper proposes cascade hybrid facial (visual) feature extraction method based on static, dynamic and key-point salient features of the face and it proves that the proposed feature extraction method is more efficient than the existing method. In this proposed method, Viola–Jones algorithm is used to detect static and dynamic features of eye, nose, lip, Scale Invariant Feature Transform (SIFT) algorithm is used to detect some stable key-point features of face. In this paper, a research on the audio-visual integration method using AND logic is also made. Furthermore, all the experiments are carried out using Artificial Neural Network (ANN) and Support Vector Machine (SVM). An accuracy of 94.90% is achieved using proposed feature extraction method. The main objective of this work is to improve the authenticity of any application using multi-modal biometric features. Adding facial features to the speech recognition improve system security because biometric features are unique and combining evidence from two modalities increases the authenticity as well as integrity of the system.

Electrocardiography (ECG) signals provides assistance to the cardiologists for identification of various cardiovascular diseases (CVD). ECG machine records the electrical activity of the heart with the assistance of electrodes placed on the patient’s body. Qualitative characterization of ECG signal reflects its sensitiveness towards distinct artifacts that resulted in low diagnostic accuracy and may lead to incorrect decision of the clinician. The artifacts are removed utilizing a robust noise estimator employing DTCWT using various threshold values and functions. The segments and intervals of ECG signals are calculated using the peak detection algorithm followed by particle swarm optimization (PSO) and the proposed optimization technique to select the best features from a considerable pool of features. Out of the 12 features, the best four features are selected using PSO and the proposed optimization technique. Comparative analysis with other feature selection methods and state-of-the-art techniques demonstrated that the proposed algorithm precisely selects principle features for handling the ECG signal and attains better classification utilizing distinctive machine learning algorithms. The obtained accuracy using our proposed optimization technique is 95.71% employing $k$-NN and neural networks. Also, 4% and 10% improvements have been observed while using $k$-NN over ANN and SVM, respectively, when the PSO technique is executed. Similarly, a 14.16% improvement is achieved while using $k$-NN and ANN over the SVM machine learning technique for the proposed optimization technique. Heart rate is calculated using the proposed estimator and optimization technique, which is in consensus with the gold standard.

Identification of skin disease has become a challenging task with the origination of various skin diseases. This paper presents a case-based reasoning (CBR) decision support system to enhance dermatological diagnosis for rural and remote communities. In this proposed work, an automated way is introduced to deal with the inconsistency problem in CBRs. This new hybrid architecture is to support the diagnosis in multiple skin diseases. The architecture used case-based reasoning terminology facilitates the medical diagnosis. Case based reasoning system retrieves the data which contains symptoms and treatment plan of the disease from the data repository by the way of matching visual contents of the image, such as shape, texture, and color descriptors. The extracted feature vector is fed into a framework to retrieve the data. The results proved using ROC curve that the proposed architecture yields high contribution to the computer-aided diagnosis of skin lesions. In experimental analysis, the system yields a specificity of 95.25% and a sensitivity of 86.77%. Our empirical evaluation has a superior retrieval and diagnosis performance when compared to the performance of other works.

Image denoising, a significant research area in the field of medical image processing, makes an effort to recover the original image from its noise corrupted image. The Pulse Coupled Neural Networks (PCNN) works well against denoising a noisy image. Generally, image denoising techniques are directly applied on the pixels. From the literature review, it is reported that denoising after frequency domain transformation is performing better since noise removal is applied over the coefficients. Motivated by this, in this paper, a new technique called the Static Thresholded Pulse Coupled Neural Network (ST-PCNN) is proposed by combining PCNN with traditional filtering or threshold shrinkage technique in Contourlet Transform domain. Four different existing PCNN architectures, such as Neuromime Structure, Intersecting Cortical Model, Unit-Linking Model and Multichannel Model are considered for comparative analysis. The filters such as Wiener, Median, Average, Gaussian and threshold shrinkage techniques such as Sure Shrink, HeurShrink, Neigh Shrink, BayesShrink are used. For noise removal, a mixture of Speckle and Gaussian noise is considered for a CT skull image. A mixture of Rician and Gaussian noise is considered for MRI brain image. A mixture of Speckle and Salt and Pepper noise is considered for a Mammogram image. The Performance Metrics such as Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), Image Quality Index (IQI), Universal Image Quality Index (UQI), Image Enhancement Filter (IEF), Structural Content (SC), Correlation Coefficient (CC), and Weighted Signal-to-Noise Ratio (WSNR) and Visual Signal-to-Noise Ratio (VSNR) are used to evaluate the performance of denoising.

This paper analyzes the effects of color quantization on standard JPEG compression. Optimized color palettes were used to quantize natural images, using dithering and chroma subsampling as optional. The resulting variations on file size and quantitative quality measures were analyzed. Preliminary results, using a small image database, show that file size suffered an average 20% increase and a concomitant loss in quality was perceived ($-$6dB PSNR, $-$0.16 SSIM and $-$9.6 Butteraugli). Color quantization present itself as an ineffective tool on JPEG compression but if necessarily imposed, on high quality compressed images, it might lead to a negligible increase in data size and quality loss. In addition dithering seems to always decrease JPEG compression ratio.