Explore chapters and articles related to this topic
Structural Health Monitoring of Existing Building Structures for Creating Green Smart Cities Using Deep Learning
Published in Amit Kumar Tyagi, Ajith Abraham, Recurrent Neural Networks, 2023
Nishant Raj Kapoor, Aman Kumar, Harish Chandra Arora, Ashok Kumar
Otsu’s method of thresholding. It is a clustering-based most commonly used method of thresholding and compatible with all types of images [137]. The principle of this method is to separate the pixels within the image into two groups. ω0 and μ0 are the ratios of the number of pixels and average gray level that are obtained from the separated image. Similarly, the background image is separated into ω1 and μ1. The total mean of the gray image is defined in the below expression: μ=ω0tμ0t+ω1tμ1t
Lane detection and fitting using the Artificial Fish Swarm Algorithm (AFSA) based on a parabolic model
Published in Lin Liu, Automotive, Mechanical and Electrical Engineering, 2017
Xiaojin Wang, Zengcai Wang, Lei Zhao
Thresholding is critical, because it allows the images to be recognised more efficiently. We use the adaptive Otsu’s method to alter intensity images into binary images. The adaptive Otsu’s method determines the threshold by splitting the histogram of the input image to minimise the variance for each of the pixel groups. Equation (2) gives the main idea of the adaptive Otsu’s method (Y. Zhang and L. Wu, 2011). σω2(t)=ω1(t)σ12(t)+ω2(t)σ22(t)
An emerging treatment technology
Published in Manish Kumar, Sanjeeb Mohapatra, Kishor Acharya, Contaminants of Emerging Concerns and Reigning Removal Technologies, 2022
S.R. Abhishek, N. Sneha, P.B.H. Karthik
Image thresholding is a basic image segmentation technique to convert a grayscale image to a binary image. This is usually done to help image processing by separating the objects or foreground pixels from background pixels. In this study, we have applied Otsu’s method for thresholding the image. This algorithm will automatically find the optimal threshold intensities. It works by searching for the threshold intensity that optimally separates the image into foreground and background classes, Figure 17.13 shows the threshold image after applying Otsu’s method.
Dance Video Motion Recognition Based on Computer Vision and Image Processing
Published in Applied Artificial Intelligence, 2023
The setting of the threshold affects the final display effect of the image. At present, the commonly used threshold setting method is Otsu method, that is, the maximum inter class variance method, also known as Otsu method. At present, Otsu method is considered as the best method to select threshold in image segmentation. Its main principle is that there is a large inter class variance between foreground and background when measuring the uniformity of gray distribution. The smaller the variance, the smaller the difference between foreground and background. Principle of Otsu method: if the total number of pixels in the image is and the gray scale range is , the number of pixels of the corresponding gray scale is for the pixel of a certain point, and the probability of occurrence of each gray scale is as follows (3).
An Empirical Review on Evaluating the Impact of Image Segmentation on the Classification Performance for Skin Lesion Detection
Published in IETE Technical Review, 2023
Lokesh Singh, Rekh Ram Janghel, Satya Prakash Sahu
Similar to classification methods, parameter tuning of most segmentation methods is approximately similar but the fine-tuning of Otsu’s binarization-ResNet50 outperformed other methods. The reason behind their superior performance is that by assuming an image of foreground and background pixels, Otsu’s method computes the optimum threshold by separating them to make intra-class variance minimum, and their inter-class variance maximum. It then iterates entire possible threshold values and computes a measure of spread for the pixel levels i.e. the pixels that either fall in foreground or background. Thus, by fine-tuning the parameters of Otsu’s method we find the threshold value where the summation of foreground and background spreads is at its minimum. The fine-tuning of ResNet50 makes the training of layers easily without enhancing the percentage of training error which helps in handling the problem of vanishing gradient by identity mapping.
Adaptively unsupervised seepage detection in tunnels from 3D point clouds
Published in Structure and Infrastructure Engineering, 2022
Kunyu Wang, Xianguo Wu, Heng Li, Fan Wang, Limao Zhang, Hongyu Chen
The Otsu method is considered one of the most effective and widely used histogram-based thresholding segmentation methods (Nakib, Oulhadj, & Siarry, 2009; Sathya & Kayalvizhi, 2011), although many other automatic thresholding methods exist, including histogram-based threshold extraction methods. The Otsu method uses an exhaustive search to select the optimal threshold (Cai, Yang, Cao, Xia, & Xu, 2014). It is considered an appropriate method to segment and extract objects from the background information of the image. In other words, the Otsu method performs well when the image consists of two or more clear objects of similar size and distribution. However, in complex environments, object segmentation and extraction will cause the main area of the image to be disturbed by changes in the lighting. Therefore, the traditional Otsu method cannot effectively distinguish defects from backgrounds. Some researchers (Fan & Lei, 2012; Ng, 2006) improved the Otsu method by modifying the weight in the cost function, which was able to separate small objects such as defects from the larger context. The method was able to find a threshold in the flat valley between the large model of the background and the small model of the defect in the histogram. These methods are satisfactory for images containing two or more visible histogram categories (not necessarily of similar shapes and sizes) and for images with visible flat regions or low grayscale distribution probability regions identified as valleys between the two categories.