Dhanekula Institute of Engineering and Technology is a private college located in Ganguru, near Vijayawada city in Andhra Pradesh. It is affiliated to JNTUK Kakinada.
The image de-fencing process is associated with the removal of fence texture and reconstructing the elegant image. The presence of occlusions like fences/barriers is obligatory and undesirable in the images captured by the hand-held cameras or low-resolution sensors. Several algorithms that are addressing the de-fencing problem are presented in the literature. Even though few reviews are available, a comprehensive survey of image de-fencing algorithms is needed since the research efforts are continuing to address this problem. The de-fencing task is usually treated as an image inpainting problem. In this review paper, the available methods are discussed lucidly. The de-fencing methods under the single image and multifocus image categories are discussed elaborately. Further, the fence removal methods for the images acquired using different cameras are analyzed. Finally, the observations concerning the exhaustive literature available and the future directions are postulated.
This study investigates the combined influence of SiO₂ nanoparticle addition and hydrogen-enriched compressed natural gas (HCNG) dual-fuel operation on the combustion, performance, and emission characteristics of a compression ignition engine fuelled with waste cooking oil biodiesel (WCO BD20). WCO biodiesel was produced using a two-stage transesterification process and blended at 20
Background Bone tissue engineering requires porous scaffold systems that combine sufficient mechanical stability, controlled degradation, and favorable biological response. Graphene oxide (GO) is a promising nanoreinforcement because of its high surface area, oxygen-containing functional groups, and ability to interact with polymeric and ceramic scaffold phases. Methods In this study, GO-reinforced hydroxyapatite/chitosan/polycaprolactone (HA/CS/PCL) biocomposite scaffolds were fabricated by solvent casting followed by freeze-drying. GO was incorporated at 0, 0.5, 1.0, 1.5, and 2.0 wt% relative to total solids. The scaffolds were evaluated for morphology, porosity, swelling, degradation, compressive strength, elastic modulus, antibacterial response, protein adsorption, and in vitro cell viability. Finite element analysis was additionally used to examine stress distribution and load-transfer behavior under quasi-static compression. Results Controlled GO incorporation improved scaffold mechanical and biological performance compared with the unreinforced scaffold. The 1.5 wt% GO scaffold showed the most balanced response, with increased compressive strength, higher elastic modulus, enhanced cell viability, improved protein adsorption, and stronger antibacterial activity. At 2.0 wt% GO, performance decreased slightly, likely because of partial nanosheet agglomeration and less uniform stress transfer. Conclusion The findings indicate that optimized GO loading can enhance HA/CS/PCL scaffold performance for bone tissue-engineering research. However, osteogenic differentiation assays, long-term degradation studies, and in vivo validation are still required before translational or load-bearing orthopedic applicability can be confirmed.
Brain disorders become more complex for individuals due to their neurological and psychological conditions. Brain disorders can be detected using various medical imaging datasets for early and accurate diagnosis, leading to effective treatment. Many existing models accurately detect affected neurological conditions and exhibit various misclassification outcomes depending on the abnormal detection rate. In this context, Adaptive Convolutional Neural Networks (ACNNs) can handle complex and large images for accurate detection and classification. In this paper, the proposed approach combines adaptive CNNs and Capsule Networks (CapsNets) to address issues in 3D medical imaging and brain disorders detection, such as brain tumors and Parkinson's disease. The proposed system was applied to one benchmark brain disorders and accurately detected the abnormal conditions. Particularly, the 3D-CNN layers extract the spatial features at various levels from high-quality MRI images. The Capsule Network enhances the features that represent the relationships among brain disorders and identifies complex patterns in the input MRI images. Experimental results achieved high accuracy of 0.99% for brain tumor detection and classification. These results indicate that the proposed approach has more potential for accurately identifying diseases.
Melanoma, Squamous Cell Carcinoma (SCC) and Basal Cell Carcinoma (BCC) are some of the most common and the most aggressive types of skin cancer and their early diagnosis is significant in improving the chances of survival of the patients. The paper is a proposal of a deep learning-based architecture that is specifically chosen to detect and classify different types of skin cancer, including melanoma, SCC, and BCC, based on dermoscopic and clinical images. The primary concept of the proposed system is a better Convolutional Neural Network (CNN) with additional convolutional layers and Global Average Pooling (GAP) to increase sensitivity, feature-discrimination, and overall classification accuracy. To enhance the diagnostic reliability the CNN is programmed to detect cancer-specific visual patterns to give a reliable detection of a possibly malignant lesion; a second stage of validation using ensemble learning is added to the CNN to further increase the accuracy of the diagnostic result. This approach will entail incorporation of the predictions of a personal CNN with finetuned transfer learning model, which will reduce the false positives and boost the certainty of the classification between melanoma, SCC and BCC. The decision is also taken depending on a mechanism of fusion between driven by confidence, which ensures that, only in cases where the same predictions are made by multiple models with high confidence, a diagnosis is made, and Explainable Artificial Intelligence (XAI) techniques, i.e., Gradient-weighted Class Activation Mapping (GradCAM) are also employed to visualize the areas of the lesion that are responsible. This improves the level of transparency and clinical interpretability of the system; therefore, raising the degree of trust between health professionals. Experimental evaluation of standardized datasets indicates that the proposed model is characterized by high sensitivity rates, robustness, and generalization rates and can be transferred to the mobile screening applications and clinical decision support systems.
