Abstract:In recent years, infrared images have been applied in more and more extensive fields and the current research of infrared image segmentation and recognition can’t satisfy the needs of practical engineering applications. The interference of various factors on infrared detectors result in the targets detected presenting the targets of low contrast, low signal-to-noise ratio (SNR) and fuzzy edges on the infrared image, thus increasing the difficulty of target detection and recognition; therefore, it is the key point to segment the target in an accurate and complete manner when it comes to infrared target detection and recognition and it has great importance and practical value to make in-depth research in this respect. Intelligent algorithms have paved a new way for infrared image segmentation. To achieve target detection, segmentation, recognition and tracking with infrared imaging infrared thermography technology mainly analyzes such features as the grayscale, location and contour information of both background and target of infrared image, segments the target from the background with the help of various tools, extracts the corresponding target features and then proceeds recognition and tracking. To seek the optimal threshold of an image can be seen as to find the optimum value of a confinement problem. As to seek the threshold requires much computation, to seek the threshold through intelligent algorithms is more accurate. This paper proposes an automatic segmentation method for infrared target image based on differential evolution (DE) algorithm and OTSU. This proposed method not only takes into consideration the grayscale information of the image, but also pays attention to the relevant information of neighborhood space to facilitate more accurate image segmentation. After determining the scope of the optimal threshold, it integrates DE’s ability of globally searching the optimal solution. This method can lower the operation time and improve the segmentation efficiency. The simulation experiment proves that this method is very effective.

Threshold Estimating in Nonlinear Diffusions Based SAR Image Decomposition

Application of DS Evidence Theory in SAR Image Edge Detection

Target Detection from SAR Images Based on Wavelet Transform De-Noise and Improved CFAR

A Hybrid Thresholding Method for Water Extraction in Envisat Image

Thresholding of Polarimetric SAR Images of Coastal Zones Based on Three-Component Decomposition and Likelihood Ratio

Research on Edge Detection for SAR Images

A Flexible Region of Interest Extraction Algorithm with Adaptive Threshold for 3-D Synthetic Aperture Radar Images

An Infrared Image Target Segmentation Based on Improved Threshold Method

Target Aspect Estimation in SAR Images via Two-Dimensional Entropic Thresholding and Radon Transform

Rayleigh-distribution based minimum error thresholding for SAR images

Research on the Depth Image Reconstruction Algorithm Using the Two-Dimensional Kaniadakis Entropy Threshold

Minimum error threshold segmentation method for SAR image based on Rayleigh distribution assumption

Change detection in SAR images based on P-M nonlinear diffusion model

Target Detection in High-Resolution SAR Image via Iterating Outliers and Recursing Saliency Depth

Adaptive Filtering for 3D SAR Data Based on Dynamic Gaussian Threshold

Target detection based on granularity computing of quotient space theory using SAR image

SAR Image Target Recognition Using Diffusion Model and Scattering Information

A New Nonlocal Iterative Trilateral Filter for SAR Images Despeckling

A Discrete Hough Transform-Based Hierarchical Statistics Detection Method for Linearly Distributed Group Targets in SAR Images

Research on SAR image denoising method based on feature extraction

Nsct Remote Sensing Image Denoising Based On Threshold Of Free Distributed Fdr