摘要
针对传统的舰船图像识别算法的鲁棒性较差的问题,研究特征匹配的舰船图像识别算法。对待识别的舰船图像做降噪和边缘强化处理,建立高斯金字塔使用SIFT方法提取预处理后的图像中的特征,并对提取的特征进行匹配。在神经网络中对匹配后特征对过滤,对过滤后的匹配特征对分类处理完成对舰船图像的识别。通过与传统图像识别算法的对比实验,证明了研究的特征匹配的舰船图像识别算法具有更佳的鲁棒性。
Aiming at the problem that the traditional ship image recognition algorithm has poor robustness,a ship image recognition algorithm based on feature matching is studied.Noise reduction and edge enhancement are applied to the recognized ship images.Gaussian pyramid is established to extract the features in the preprocessed images using SIFT method,and the extracted features are matched.In the neural network,the matched features are filtered,and the filtered matched features are classified to complete the recognition of ship images.Compared with the traditional image recognition algorithm,the proposed algorithm is proved to be more robust.
作者
李海燕
LI Hai-yan(Yantai Vocational College,Yantai 264670,China)
出处
《舰船科学技术》
北大核心
2020年第8期46-48,共3页
Ship Science and Technology
基金
山东省高等学校科技计划项目(J18KB172)
作者简介
李海燕(1973-),女,硕士,讲师,主要从事计算机软件设计。
