Effect of Gaussian filtered images on Mask RCNN in detection and segmentation of potholes in smart cities

Abstract

<abstract> <p>Accidents have contributed a lot to the loss of lives of motorists and serious damage to vehicles around the globe. Potholes are the major cause of these accidents. It is very important to build a model that will help in recognizing these potholes on vehicles. Several object detection models based on deep learning and computer vision were developed to detect these potholes. It is very important to develop a lightweight model with high accuracy and detection speed. In this study, we employed a Mask RCNN model with ResNet-50 and MobileNetv1 as the backbone to improve detection, and also compared the performance of the proposed Mask RCNN based on original training images and the images that were filtered using a Gaussian smoothing filter. It was observed that the ResNet trained on Gaussian filtered images outperformed all the employed models.</p> </abstract>

Enhanced Lung Nodule Segmentation using Dung Beetle Optimization based LNS-DualMAGNet Model

“…To lessen Gauss noise at specific spots in the lung nodule image, a specialised filter called the GSFilter is applied. Poisson is created, however, by the statistical properties of electromagnetic waves, such as gamma or X-rays forms a convolution matrix using values from the Gaussian distribution to depict the approximate Gaussian distribution (or "bell-shaped" hump) using a convolutional kernel [26]. We can stop the kernel at three standard deviations from the average since, in reality, the distribution of values is effectively zero at this point.…”

STD-Detector: spatial-to-depth feature-enhanced detection method for the surface defect detection of strip steel

“…The current object detection model in the surface defect detection of steel strips field can be broadly categorized into three main types: two-stage object detection model, one-stage object detection model, and transformer-based object detection model called DETR 5 – 7 Examples of two-stage object detection model include region-based convolutional neural network (RCNN), 8 Fast RCNN, 9 Faster RCNN, 10 – 12 and Mask RCNN 13 , 14 . One-stage object detection model encompasses the you only look once (YOLO) series 15 , 16 and single shot multibox detector (SSD) 17 , 18 .…”

“…[5][6][7] Examples of two-stage object detection model include region-based convolutional neural network (RCNN), 8 Fast RCNN, 9 Faster RCNN, [10][11][12] and Mask RCNN. 13,14 One-stage object detection model encompasses the you only look once (YOLO) series 15,16 and single shot multibox detector (SSD). 17,18 Two-stage object detection model generates candidate boxes, extracts feature from the region contents within those boxes, and then performs target regression.…”