Melanoma or malignant melanoma is a type of skin cancer that develops when melanocyte cells, damaged by excessive exposure to harmful UV radiations, start to grow out of control. Though less common than some other kinds of skin cancers, it is more dangerous because it rapidly metastasizes if not diagnosed and treated at an early stage. The distinction between benign and melanocytic lesions could at times be perplexing, but the manifestations of the disease could fairly be distinguished by a skilled study of its histopathological and clinical features. In recent years, deep convolutional neural networks (DCNNs) have succeeded in achieving more encouraging results yet faster and computationally effective systems for detection of the fatal disease are the need of the hour. This paper presents a deep learning-based ‘You Only Look Once (YOLO)’ algorithm, which is based on the application of DCNNs to detect melanoma from dermoscopic and digital images and offer faster and more precise output as compared to conventional CNNs. In terms with the location of the identified object in the cell, this network predicts the bounding box of the detected object and the class confidence score. The highlight of the paper, however, lies in its infusion of certain resourceful concepts like two phase segmentation done by a combination of the graph theory using minimal spanning tree concept and L-type fuzzy number based approximations and mathematical extraction of the actual affected area of the lesion region during feature extraction process. Experimented on a total of 20250 images from three publicly accessible datasets—PH2, International Symposium on Biomedical Imaging (ISBI) 2017 and The International Skin Imaging Collaboration (ISIC) 2019, encouraging results have been obtained. It achieved a Jac score of 79.84% on ISIC 2019 dataset and 86.99% and 88.64% on ISBI 2017 and PH2 datasets, respectively. Upon comparison of the pre-defined parameters with recent works in this area yielded comparatively superior output in most cases.

Information hiding or data hiding, also known as watermarking, has become a part and parcel of covert communication and copyright protection. Maximizing watermark payload is a major challenge for watermark researchers. To overcome this issue, we have proposed a new color image watermarking technique, using residue number system (RNS). RNS refers to a large integer using a set of smaller integers which relies on the Chinese remainder theorem of modular arithmetic for its operation. The proposed method takes pixel values from three watermark images and embeds them into the main cover image. Experimental results presented in this paper shows that the watermark can be successfully embedded and extracted from an image, without distorting the original image using the proposed technique. The high peak signal to noise ratio (PSNR) and payload values claims the robustness of the proposed method.

Salt and pepper noise removal is extremely challenging task when the noise density is very high in the corrupted image. Usually the existing filters are not capable of removing this noise above 70% density. This paper proposes a Sequentially Combined Mean-Median Filter (SCMMF) based on combination of arithmetic mean and median technique. A binary flag image is generated with the help of identified noisy pixels which are first replaced by a value calculated by plus (+) operation with nearest pixels intensities. The remaining noisy pixels are then replaced by neighbourhood mean operations. Finally the image quality is enhanced by the median of the corresponding window. Border operation is also incorporated to preserve the image size. Experimental results show that SCMMF is outstanding from visual qualitative judgments for a number of standard images. Moreover, this combined filter outperform over the existing with respect to MSE, PSNR, and SSIM comparison even at 92% noise density level.

Non-Functional Requirements (NFRs) such as performance, security, reliability, and usability significantly influence the overall quality of software systems. However, traditional NFR engineering approaches commonly treat these requirements as binary constraints, i.e., either fully satisfied or not, which limits systematic trade-off reasoning even in conventional single-system software architectures. This limitation becomes critical in complex, dynamic domains such as IoT and cyber-physical systems, where multiple interdependent NFRs must coexist without degrading system performance. To address this challenge, this paper proposes a domain-specific framework for modeling and optimizing priority-based NFRs beyond binary constraints. The framework integrates standard well-recognized quality models like ISO/IEC 25010 and FURPS+, to identify domain-specific NFRs, which are then modeled as soft goals with graded degrees of satisfaction. Priority weights, utility-based reasoning, and multi-objective optimization are integrated to manage conflicts and trade-offs among competing NFRs while preserving system stability. Unlike traditional approaches that enforce a rigidly defined subset of NFRs, the proposed framework enables the combination of multiple NFRs proportionally according to their priority and contextual relevance. The framework is applicable to diverse software-intensive systems including IoT, cloud platforms, healthcare systems, and cyber-physical environments. This work contributes a systematic, scalable, and adaptive approach to proactive NFR management suitable for modern software systems. The concrete research problem addressed is the failure of binary NFR handling in dynamic adaptation settings, leading to over-constrained designs or degraded quality. The framework guarantees improved global utility (15-25% in simulations) and adaptation success (>90%) while maintaining performance thresholds, with low runtime overhead (~O(n²) for n NFRs).