基于计算机技术的加密纸张图像识别模型构建
Construction of an Encrypted Paper Image Recognition Model Based on Computer Technology
罗钢 1许珂乐1
作者信息
- 1. 咸阳职业技术学院,陕西 咸阳,712000
- 折叠
摘要
为实现造纸企业对加密纸张图像的自动化识别,提高数字图像识别准确率,通过伽马变换算法提升数字图像的对比度并对其实施边缘检测处理,进而实现原始纸张加密图像的外观增强,并获取用于识别加密纸张的目标图像.在此基础上,基于O-FAST角点检测技术实现目标图像与待测图像之间的近距离匹配,进而达到针对加密纸张图像的识别效果.为验证该方法的有效性,专门组建了一个图像识别匹配系统用于统计图像匹配效果.经试验研究发现,加密纸张图像识别模型在各匹配项目中的准确率均在 95%以上.
Abstract
To achieve automated recognition of encrypted paper images and improve the accuracy of digital image recognition in papermaking enterprises.This study uses the gamma transform algorithm to enhance the contrast of digital images and applies edge detection processing,thereby enhancing the appearance of encrypted images of original paper and obtaining target images for identifying encrypted paper.On this basis,based on O-FAST point detection technology,close range matching between the target image and the test image is achieved,thereby achieving recognition effect for encrypted paper images.To verify the effectiveness of this method,an image recognition matching system was specifically developed in this study to statistically evaluate the image matching performance.Through experimental research,it was found that the accuracy of the encrypted paper image recognition model in all matching items is above 95%.
关键词
伽马变换/O-FAST角点检测/特征点匹配/图像识别Key words
gamma transform/O-FAST point detection/feature point matching/image recognition引用本文复制引用
基金项目
咸阳职业技术学院科研基金项目(2021KJB10)
咸阳职业技术学院科研项目(2021KJB05)
陕西省教育厅2023年度陕西高等职业教育教学改革研究项目(23GZ025)
出版年
2024
NETL
NSTL
万方数据