Research On Mineral Microscopic Image Enhancement Technology

Mineral microscopic image analysis has become one of the important methods for evaluating mineral quality and properties.By virtue of ultra-high resolution and large depth of field,scanning electron microscope(SEM)is used to observe the surface of samples with uneven structure.Therefore,it has been widely applied in mineral image acquisition and analysis.However,due to the inherent properties of mineral samples,the microscopic images acquired by SEM are blurry and characterized by low contrast,which seriously affect the accuracy of the mineral micro-structure analysis.Consequently,it is meaningful to enhance microscopic image and improve its visuality.Driven by the practical requirement,this thesis investigates a serious of image enhancement methods to improve the quality of distorted mineral images obtained by SEM.The main contents are as follows:To deal with the mineral microscopic images of low contrast,this thesis conducted research based on two cases.In terms of images with explicit details in the bright and dark regions,the enhancement method is implemented in the following steps.Firstly,the images are processed by double platform histogram equalization;Secondly,the local details are enhanced through the connected domain equalization based on recursive hierarchy.Specifically,this method divides the output image in a recursive way and localizes the output components with the connected components of each layer.Finally,the processing results are fused to get the final image.The experiment results show that this method has a significant effect on improving image contrast and enhancing details of bright and dark regions,and achieved high information entropy and average gradient value in objective evaluation.At the same time,a contrast enhancement method based on gradient transform has been presented for images with interesting information only in the salient target region.This method enhances the image gradient by setting the threshold as the enhancement target,and reconstructs the enhanced gradient field to obtain the enhanced image of the salient region.The simulation results show that the low contrast images enhanced by the method are clearer in target region and with more obvious boundaries between background and object.For the mineral samples with sharp edges,the output of the secondary electrons at their sharp edges is very high,which generates high brightness at the edges and specifically poor brightness in the flat region of the images.To tackle this problem,anbrightness unevenness correction method of adaptive gamma function is proposed,which uses the bilateral filter to extract the brightness component of the image.This function could adaptively adjust the gamma value according to the different brightness of different regions.As a result,the brightness of the target edge is suppressed while details in the flat are effectively highlighted.Compared with other methods such as global gamma correction,the method can achieve a better improvement of bright and dark anomalies in the image.