基于图像检测的硬币面值识别方法

An Identification Method of Coins Denomination Based on Image Detection

为了实现同时对多个硬币面值的准确识别,文中提出一种基于图像检测的硬币面值识别方法。首先逐步采用中值滤波、阈值分割以及形态学操作对灰度图像中的硬币进行ROI提取,实现硬币的粗定位;然后以ROI的外接最小矩形作为研究区域从原图提取图像,并由RGB颜色空间转化到HSV颜色空间,通过颜色特征的不同识别出5角硬币;最后采用SSIM图像质量评价的方法将剩余ROI中的1角硬币与1元硬币检测识别。结果表明,该硬币面值识别方法在不需要精确获得硬币大小尺寸的情况下,可以同时对多个硬币进行准确识别。

In order to realize accurate recognition of multiple coin denominations at the same time,a recognition method of coins denomination based on image detection was proposed in this paper.Firstly,median filtering,threshold segmentation and morphological operations were gradually used to extract ROI of coins in grayscale image to realize rough coins positioning.Then,the smallest rectangles outside ROI were taken as the research area to extract the images from the original image.The extracted images were transformed from RGB color space to HSV color space,so the 5-dime coins could be identified through the difference of color features.Finally,SSIM image quality evaluation method was used to detect and identify the 1-dime coins and 1-yuan coins in the remaining ROI.The result showed that the identification method of coins denomination could accurately identify multiple coins at the same time without accurately obtaining the coin size.