基于算力消耗与灵敏度的图像相似度算法对比  

Comparison of Image Similarity Algorithms Based on Computing Power Consumption and Sensitivity

作　　者：关达伟 林欣平 张昊一 GUAN Dawei;LIN Xinping;ZHANG Haoyi(Beijing Jiaotong University(Weihai Campus),Weihai Shandong 264400)

机构地区：[1]北京交通大学(威海校区),山东威海264400

出　　处：《软件》2024年第5期46-48,共3页Software

摘　　要：本文致力于研究在工业场景下分析图像相似度算法。考虑灵敏度和算力需求,通过横向比较不同算法的计算时间和灵敏度,在随机和梯度数据集下评估性能,以找出满足实验预期的最优化算法,即为找出不合格产品的最优化算法。This paper is devoted to the study of image similarity analysis algorithms in industrial scenes.Taking into account sensitivity and computing power requirements,performance is evaluated on random and gradient datasets by comparing the computation time and sensitivity of different algorithms horizontally to find the optimal algorithm that meets the experimental expectation,it isthe optimal algorithm to find unqualified products.

关 键 词：工业产品 相似度检测 算法灵敏度分析 

分 类 号：TP311.53[自动化与计算机技术—计算机软件与理论]