Image segmentation plays a pivotal role in computer vision applications, particularly within the agriculture sector, where it plays a critical role in applications such as the early diagnosis of plant diseases.
Traditional color image thresholding methods face challenges in determining optimal thresholds, especially for large number of thresholds, that negatively impacts the accuracy of the segmentation process.
In this paper, we propose a Newton Raphson-Based Optimizer (NRBO), a novel algorithm for multi-level
threshold image segmentation using Kapur's entropy as the objective function. The proposed NRBO incorporates two key components: the Newton Raphson Search Rule (NRSR) to enhance convergence speed
and the Trap Avoidance Operator (TAO) to prevent local optima. These components improve the algorithm exploration and exploitation capabilities. The proposed NRBO was applied to segment ten images of tomato leaf diseases. The performance of the proposed NRBO was compared against other optimization algorithms, including Reptile Search Algorithm, Ant Lion Optimizer, Atom Search Optimizer, and Black Widow Optimizer. Experimental results demonstrated that the proposed NRBO consistently achieved superior performance across evaluation metrics, including Feature Similarity Index (FSIM), Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index (SSIM), and CPU time, while maintaining competitive computational efficiency. For instance, NRBO achieved the highest FSIM, PSNR, and SSIM values for the majority of tested images, indicating its robustness in handling multi-level threshold segmentation under various conditions. The results highlight the effectiveness of NRBO in addressing image segmentation challenges, making it a promising solution for diagnosing plant diseases.