一种基于忆阻交叉阵列的自适应三高斯模型及其在图像增强中的应用

发布时间：2019-02-19 17:16

【摘要】：在视觉图像处理中,可用三高斯模型来模拟视网膜神经节细胞的感受野,这可以在一定程度上对图像信息,比如图像的边缘、细节等信息进行增强.但是在对大量的图像进行处理时,为了达到比较理想的效果,就需要人为地来改变模型中的相关参数,这是一个十分耗时的过程.基于此,本文提出一种基于忆阻交叉阵列的自适应三高斯模型.这种模型是在传统三高斯模型的基础上,根据所需处理图像的局部特征,利用忆阻交叉阵列的特性动态地改变模型参数,以达到对局部图像最优增强的目的,从而使整幅图像的增强效果更好.首先,根据图像的局部亮度信息来确定忆阻器所需施加的脉冲电压的极性以及宽度;然后,根据所得忆阻值得到对应模型中参数的值;最后,可以得到局部增强模板,从而实现增强.本文分别选取了彩色和灰度图像进行了测试,定性和定量实验结果均表明,这种改进的三高斯模型不仅能够对图像边缘进行有效的增强,而且还可以极大地提高图像的对比度和清晰度,为忆阻器在图像处理方面的应用提供了新方向.
[Abstract]:In visual image processing, three Gao Si models can be used to simulate the receptive field of retinal ganglion cells, which can enhance the image information to a certain extent, such as image edge, details and so on. However, when processing a large number of images, it is necessary to change the relevant parameters of the model artificially in order to achieve a better effect, which is a very time-consuming process. Based on this, an adaptive three Gao Si model based on amnesia crossover array is proposed. This model is based on the traditional three Gao Si model, according to the local characteristics of the image to be processed, using the characteristics of the amnesia crossover array to dynamically change the parameters of the model, in order to achieve the purpose of optimal enhancement of the local image. Thus, the enhancement effect of the whole image is better. Firstly, the polarity and width of the pulse voltage applied to the resistor are determined according to the local luminance information of the image, and then the corresponding parameters in the model are obtained according to the obtained values. Finally, local enhancement templates can be obtained to achieve enhancement. The qualitative and quantitative experimental results show that the improved three-Gao Si model can not only enhance the edge of the image effectively. Moreover, it can greatly improve the contrast and sharpness of the image, which provides a new direction for the application of the resistive device in image processing.
【作者单位】：西南大学电子信息工程学院非线性电路与智能信息处理重庆市重点实验室;
【基金】：国家自然科学基金(批准号:61571372,61672436,61372139) 中央高校基本科研业务费专项资金(批准号:XDJK2016A001,XDJK2014A009) 新世纪优秀人才支持计划(批准号:教技函[2013]47号)资助的课题~~
【分类号】：TP391.41

【相似文献】