基于Fisher准则函数的图像阈值分割模型构建

Image threshold segmentation model based on Fisher criterion function

图像分割是图像处理技术中较为热门的研究对象之一,目前图像阈值分割成为图像分割技术中的主要方式之一。但传统的图像阈值分割模型阈值计算误差较大,导致在多目标的情况下无法获得完整的图像分割结果。因此,构建基于Fisher准则函数的图像阈值分割模型。设计Fisher准则算法,将原始图像平滑处理并得到阈值计算标准。根据Fisher准则与分割阈值计算标准,完成图像边缘识别,计算图像边缘的灰度权重,得到图像分割阈值。使用人工蜂群算法在多个分割阈值计算结果中,得到最优分割阈值,并使用此分割阈值完整图像分割过程。构建实验环节,通过与其他两种模型对比可知,此模型可对多个目标进行分割,同时此模式的分割阈值计算误差较小。综合上述实验结果可知,此模型的使用效果优于目前使用的分割模型,且此模型的综合性能较高。

Image segmentation is one of the most popular research objects in image processing technology.At present,image threshold segmentation has become one of the main methods in image segmentation technology.However,the traditional image threshold segmentation model has a large error in threshold calculation,which leads to the inability to obtain complete image segmentation results in the case of multiple targets.Therefore,an image threshold segmentation model based on Fisher criterion function is constructed.Fisher criterion algorithm is designed to smooth the original image and get the threshold calculation standard.According to the Fisher criterion and the segmentation threshold calculation standard,the image edge recognition is completed,the gray weight of image edge is calculated,and the image segmentation threshold is obtained.The artificial bee colony algorithm is used to get the optimal segmentation threshold from the calculation results of multiple segmentation thresholds.The segmentation threshold is used to complete the image segmentation process.In comparison with the other two models,this model can segment multiple targets,and its calculation error of segmentation threshold is small.The above experimental results indicate that the effect of this model is better than that of the current segmentation model,and the comprehensive performance of this model is higher.