摘要
传统船舶视频图像类型识别算法在图像类型识别计算过程中,存在视频图像高频尺度空间像素子带系数计算精度较低,导致图像画面差异辨识度降低,无法高精度对比图像特征参量的问题。因此,提出计算机大数据船舶视频图像类型辩识分析。通过神经网络算法,建立视频图像特征的神经网络模型;根据特征模型推导建立辨识模型。在辨识模型的基础上,通过导入大数据高频子带尺度算法,对图像特征尺度空间内的高频子带系数进行优化计算,从而提升视频图像类型的辨识精度。通过实验数据对比表明:提出的视频图像辨识算法,能够有效改善图像类型识别环境,提升图像类型辨识精度,解决了传统辨识算法辨识精度低的问题。
In the process of image type recognition calculation, the traditional ship video image type recognition algorithm has low accuracy in calculating the pixel subband coefficient of high frequency scale space of video image, which leads to the decrease of image differential recognition. The problem of unable to compare image feature parameters with high precision. Therefore, the identification and analysis of ship video image type of computer big data is proposed. The neural network model of video image features is established by neural network algorithm, and the identification model is derived according to the feature model. On the basis of the identification model, the high frequency subband scale algorithm of big data is introduced to optimize the calculation of the high frequency subband coefficients in the image feature scale space. From And improve the accuracy of video image type identification. The experimental data show that the proposed video image recognition algorithm can effectively improve the image type recognition environment, improve the image type identification accuracy, and solve the problem of low recognition accuracy of traditional identification algorithm.
作者
高洁
GAO Jie(University of Electronic Science and Technology of China,Chengdu 611731,China;Hebei Women's Vocational College,Shijiazhuang 050091,China)
出处
《舰船科学技术》
北大核心
2021年第2期82-84,共3页
Ship Science and Technology
关键词
大数据
船舶
视频图像
类型辨识
Big Data
Ship
Video Image
Type Identification
