Spatial Context Energy Curve-Based Multilevel 3-D Otsu Algorithm for Image Segmentation

Bhandari, Ashish Kumar; Singh, Anurag; Kumar, Immadisetty Vinod

doi:10.1109/tsmc.2019.2916876

Cited by 26 publications

(13 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Among various methods of thresholding [1,3,16,22,49,50], MCE has been quite popular and is used to select the optimal threshold. It reduces the cross entropy between the segmented and the original test image.…”

Section: Multilevel Minimum Cross Entropymentioning

confidence: 99%

See 1 more Smart Citation

An improved segmentation technique for multilevel thresholding of crop image using cuckoo search algorithm based on recursive minimum cross entropy

Kumar

Vishwakarma

et al. 2022

IET Signal Processing

View full text Add to dashboard Cite

Crop image segmentation is widely used for the analysis of crops. A wide variety of crops are present in the agriculture field, which varies in intensity and complex backgrounds. The thresholding method based on entropy is quite popular for the segmentation of an image. Among all, minimum cross entropy has been widely used. However, the complexity of computation increases when it is used for multilevel thresholding (MLT). Recursive minimum cross entropy is used to resolve the complexity of computation, and cuckoo search (CS) using Levy flight is used to find the optimal threshold for this objective function. Because real-time applications require less processing time while maintaining high performance, which is validated by the CS algorithm using recursive minimum cross entropy (R-MCE-CS) without constraint. The proposed method uses one constraint based on the structural similarity index (SSIM), which leads to an increment in the accuracy for a higher level of thresholding. The accuracy of the proposed method has been tested over 10 crop images with complex backgrounds and high dimensions of colour intensity space. The outcome of the proposed technique has been compared with five algorithms such as wind-driven optimisation (WDO), bacterial foraging optimisation (BFO), differential evolution (DE), artificial bee colony (ABC), and firefly algorithm (FFA). The result shows that the proposed method gives the most promising result, and the accuracy is also improved.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

Section: Multilevel Minimum Cross Entropymentioning

confidence: 99%

“…for separation of an image. The 2‐level thresholding consists of one valley between two peaks, whereas several peaks and valleys are found in MLT [16]. The MLT approach is more complex but the significance is increasing by the day [17, 18].…”

Section: Introductionmentioning

confidence: 99%

An improved segmentation technique for multilevel thresholding of crop image using cuckoo search algorithm based on recursive minimum cross entropy

Kumar

Vishwakarma

et al. 2022

IET Signal Processing

View full text Add to dashboard Cite

show abstract

“…The modified Hopfield neural network (HNN) architecture [8,37] is used to derive the energy curve for an image. The HNN architecture consists of a set of neurons interconnected by weighting functions.…”

Section: Energy Curve For Imagementioning

confidence: 99%

Spatial Information Computation-Based Low Contrast Image Enhancement

Srinivas

Bhandari

2021

Circuits Syst Signal Process

Self Cite

View full text Add to dashboard Cite

This paper proposes a novel method of image enhancement via energy curve equalization (ECE). Histogram equalization (HE) appears to be the most straightforward method for image contrast enhancement. Many modifications are already recommended to overcome the confines of HE. The computation of the histogram does not consider the spatial correlations among the surrounding pixels. The current method utilizes the energy curve as a substitute for the histogram of an image. In contrast to the histogram, this approach covers the spatial context information of neighboring pixels. The modified Hopfield neural network (HNN) architecture is employed to compute an image's energy curve. It has peaks and valleys and appears smoother than the image histogram. The average value of the mean and median of the energy curve is used as the plateau limit to control the enhancement rate. The clipped energy curve is partitioned into three individual regions based on the standard deviation value. The final transformation function is utilized to remap the pixel intensity values and is formulated by integrating the individual portions of equalized energy curves. The simulation results confirm the projected technique's effectiveness compared to conventional HE-based methods with and without plateau limit.

show abstract

“…Similarly, based on the Otsu method, Chou et al [ 5 ] proposed a region-based image segmentation algorithm, and Wen et al [ 6 ] proposed a curvelet transform method; however, the former was poorly adaptive, and the latter was sensitive to noise. Bhandari et al [ 7 ] suggested that the multilevel 3-D Otsu with spatial context performed better than the histogram-based Otsu method in segmentation. The text image binarization algorithm based on the precipitation model was proposed by Oh et al [ 8 ].…”

Section: Introductionmentioning

confidence: 99%

Spectrum decomposition in Gaussian scale space for uneven illumination image binarization

et al. 2021

View full text Add to dashboard Cite

Although most images in industrial applications have fewer targets and simple image backgrounds, binarization is still a challenging task, and the corresponding results are usually unsatisfactory because of uneven illumination interference. In order to efficiently threshold images with nonuniform illumination, this paper proposes an efficient global binarization algorithm that estimates the inhomogeneous background surface of the original image constructed from the first k leading principal components in the Gaussian scale space (GSS). Then, we use the difference operator to extract the distinct foreground of the original image in which the interference of uneven illumination is effectively eliminated. Finally, the image can be effortlessly binarized by an existing global thresholding algorithm such as the Otsu method. In order to qualitatively and quantitatively verify the segmentation performance of the presented scheme, experiments were performed on a dataset collected from a nonuniform illumination environment. Compared with classical binarization methods, in some metrics, the experimental results demonstrate the effectiveness of the introduced algorithm in providing promising binarization outcomes and low computational costs.

show abstract

Spatial Context Energy Curve-Based Multilevel 3-D Otsu Algorithm for Image Segmentation

Cited by 26 publications

References 28 publications

An improved segmentation technique for multilevel thresholding of crop image using cuckoo search algorithm based on recursive minimum cross entropy

An improved segmentation technique for multilevel thresholding of crop image using cuckoo search algorithm based on recursive minimum cross entropy

Spatial Information Computation-Based Low Contrast Image Enhancement

Spectrum decomposition in Gaussian scale space for uneven illumination image binarization

Contact Info

Product

Resources

About