Search

Driver drowsiness and fatigue are leading causes of road accidents, particularly due to drivers having to operate vehicles for long hours under challenging physical and adverse weather conditions. Despite extensive research on subjective, physiological, and vehicle-based detection methods, recent reports still highlight drowsiness and fatigue as significant contributors to accidents and injuries on highways. However, existing drowsiness detection models are often intrusive or expensive. To address these limitations, we propose a low-cost, non-intrusive drowsiness detection model based on facial features, leveraging principles of image processing and machine learning. Our research involves a multi-step approach. Initially, image frames are extracted and subjected to image processing techniques. The images are subsequently passed through a Dlib classifier to detect and localize 68 facial landmarks, including the mouth and eyes. From the localized eye landmarks, salient features such as eye aspect ratio (EAR), pupil circularity (PUC), and eyelid closure over the eye pupil are extracted. To classify the driver's state, these collective features are inputted into various machine learning classifiers. We conducted experiments using the National Tsing Hua University Computer Vision Lab dataset, implementing the proposed approach in the Python OpenCV environment. Notably, the KNN algorithm demonstrated the highest accuracy when the test data was evaluated, yielding a paramount accuracy of 78% on the NTHU DDD dataset. In conclusion, the non-intrusive facial features-based driver drowsiness detection model offers a cost-effective solution to address the pervasive problem of drowsiness-related accidents. The promising results obtained through the utilization of machine learning techniques underscore the potential for practical implementation. [ABSTRACT FROM AUTHOR]

Diabetic retinopathy is a consequence of diabetes which is brought on by high blood sugar levels harming the retina. Patients with diabetic retinopathy start to notice fuzzy vision, color blindness, and variances in vision. Floating dots in vision, night vision deterioration, and dark regions in the visual field when diabetic retinopathy advances to more severe stages. If undetected and mistreated, it can lead to blindness. However, the progression takes years for diabetic retinopathy to endanger your vision. The earliest apparent symptoms of diabetic retinopathy are micro aneurysms, which manifest as tiny red spots on the retina. Other diabetic retinopathy symptoms include hemorrhages, which are red lesions brought on by the rupture of tiny blood vessels in the retina's deeper layers. Exudates are another type of typical feature of diabetic retinopathy. These lesions are yellow-white and are brought on by plasma leakage from capillaries. Numerous physical examinations, including those for visual acuity, pupil dilation, and optical coherence tomography, can be performed to find diabetic retinopathy, but they take time as well as are costly, and have negative effects on the patients. This paper presents a novel image processing technique and algorithm for the enhancement of eye fundus images for the effective diagnosis of diabetic retinopathy. This paper also highlights the comparison between the effectiveness of different image processing techniques including negative transform, log transform, and contrast stretching techniques. It further compares the various contrast stretching methods including Contrast Limited Adaptive Histogram Equalization (CLAHE), Histogram Equalization (HE), and Adaptive histogram equalization (AHE). The paper concludes with the most effective image processing technique for the diagnosis of diabetic retinopathy that can help to treat the disease in its earlier stages. [ABSTRACT FROM AUTHOR]

The brain is a vital and complex organ of the human system. It is responsible for controlling the overall activity of the human body. Normal functionality of the organs depends on the healthy brain and any disorder in the brain may cause a critical and life-threatening condition. There are several disorders related to the brain like Alzheimer's, Tumors, Cancer, Epilepsy, and Stroke. The severity of the brain disorder depends on the region and its intensity. Among all these disorder brain Tumors is the most common which may occur in any age and gender. There are several modalities used to acquire the image of the brain to diagnose a brain tumor but identifying the tumor from the image requires profound knowledge and expertise. Physically identification of the tumor from the radiologic image is a common practice of medical professionals; however, the procedure has the probability of human error. Computer-aided diagnoses of Brain Tumors assist the medical professional to estimate the proper region of the tumor in an image in a better way. Several approaches have been adopted by the researcher to identify the brain tumor from a radiologic image. This article proposed a MATLAB-based graphical user interface to assist medical professionals to estimate Brain Tumors from radiologic Images. The algorithm is based on the acquisition of the radiological image, preprocessing, Image segmentation, and algorithm implementation for Tumor detection. Further, the Convolutional Neural Network techniques have been applied to compare the accuracy of the six different pre-trained models. Significant results have been achieved via the proposed algorithm. The algorithm reduces the process time and rate of human error cost-effectively. The ResNET50 Approach-1 is identified as the better approach among the six approaches with a Training and Validation Accuracy of 99 % and Test Accuracy of 93%. [ABSTRACT FROM AUTHOR]

Brain tumor segmentation plays a vital role in medical image analysis, aiding in accurate diagnosis and treatment planning. This research article proposes a novel approach for brain tumor segmentation utilizing the K-Means clustering algorithm and Fuzzy C-Means (FCM) clustering technique. The K-means segmentation is initially used to cluster the brain image into distinct regions based on intensity values. Subsequently, the FCM algorithm is applied to further refine the segmentation results by considering the fuzzy memberships of pixels within each cluster. The combination of K-means segmentation and FCM algorithm improved the tumor detection accuracy and also reduced computational complexity compared to traditional methods. The objective is to enhance the accuracy and efficiency of tumor segmentation in magnetic resonance imaging (MRI) scans. [ABSTRACT FROM AUTHOR]