Hybrid Adaptive Crayfish Optimization with Differential Evolution for Color Multi-Threshold Image Segmentation

Biomimetics, Journal Year: 2025, Volume and Issue: 10(4), P. 218 - 218

Published: April 2, 2025

To better address the issue of multi-threshold image segmentation, this paper proposes a hybrid adaptive crayfish optimization algorithm with differential evolution for color segmentation (ACOADE). Due to insufficient convergence ability in later stages, it is challenging find more optimal solution optimization. ACOADE optimizes maximum foraging quantity parameter p and introduces an adjustment strategy enhance randomness algorithm. Furthermore, core formula (DE) incorporated balance ACOADE’s exploration exploitation capabilities better. validate performance ACOADE, IEEE CEC2020 test function was selected experimentation, eight other algorithms were chosen comparison. verify effectiveness threshold Kapur entropy method Otsu used as objective functions compared algorithms. Subsequently, peak signal-to-noise ratio (PSNR), feature similarity index measure (FSIM), structural (SSIM), Wilcoxon employed evaluate quality segmented images. The results indicated that exhibited significant advantages terms value, metrics, convergence, robustness.

Language: Английский

Multi-scale parallel gated local feature transformer

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 5, 2025

Visual Simultaneous Localization and Mapping (VSLAM) is a crucial technology for autonomous mobile vision robots. However, existing methods often suffer from low localization accuracy poor robustness in scenarios with significant scale variations low-texture environments, primarily due to insufficient feature extraction reduced matching precision. To address these challenges, this paper proposes an improved multi-scale local algorithm based on LoFTR, named MSpGLoFTR. First, we introduce Multi-Scale Local Attention Module (MSLAM), which achieves fusion resolution alignment through window partitioning shared multi-layer perceptron (MLP). Second, Parallel designed capture features across various scales, enhancing the model's adaptability large-scale highly similar pixel regions. Finally, Gated Convolutional Network (GCN) mechanism incorporated dynamically adjust weights, emphasizing key while suppressing background noise, thereby further improving precision robustness. Experimental results demonstrate that MSpGLoFTR outperforms LoFTR terms of precision, relative pose estimation performance, complex scenarios. Notably, it excels environments illumination changes, variations, viewpoint shifts. This makes efficient robust solution tasks.

Language: Английский

FEBE-Net: Feature Exploration Attention and Boundary Enhancement Refinement Transformer Network for Bladder Tumor Segmentation

Mathematics, Journal Year: 2024, Volume and Issue: 12(22), P. 3580 - 3580

Published: Nov. 15, 2024

The automatic and accurate segmentation of bladder tumors is a key step in assisting urologists diagnosis analysis. At present, existing Transformer-based methods have limited ability to restore local detail features insufficient boundary capabilities. We propose FEBE-Net, which aims effectively capture global remote semantic features, preserve more information, provide clearer precise boundaries. Specifically, first, we use PVT v2 backbone learn multi-scale feature representations adapt changes tumor size shape. Secondly, new exploration attention module (FEA) fully explore the potential information shallow extracted by backbone, eliminate noise, supplement missing fine-grained details for subsequent decoding stages. same time, enhancement refinement (BER), generates high-quality clues through detection operators help decoder refine adjust final predicted map. Then, efficient self-attention calibration (ESCD), which, with provided BER module, gradually recovers contextual from high-level after low-level attention. Extensive experiments on cystoscopy dataset BtAMU five colonoscopy datasets shown that FEBE-Net outperforms 11 state-of-the-art (SOTA) networks performance, higher accuracy, stronger robust stability, generalization ability.

Language: Английский