A new method for image thresholding has been developed to improve accuracy and speed in segmenting images with low signal-to-noise ratios. By combining grayscale-gradient and traditional histograms, noise can be removed more effectively, leading to better results. This method simplifies the process of finding the optimal threshold, making it faster and more robust compared to traditional techniques. The experimental results show that this improved method is more accurate and efficient, especially for images with low signal-to-noise ratios.