Text in natural scenes is often affected by complex text background, irregular text arrangement and other factors, resulting in a significant increase in the difficulty of text detection. In order to further improve the accuracy of text detection, an improved text detection method AC-DBNet for complex natural scenes is proposed based on the segment-based detection technology DBNet. The main points of improvement are to strengthen the feature extraction capability of DBNet network model, add the following extraction module CEM module before the feature map fusion, increase the receptive file through hole convolution, and add the attention guidance module AM to enhance the detection of small target text area, and improve the accuracy of the detection network through these improvements. The accuracy of the improved algorithm in the test set reached 92.1%, which was 0.7% higher than that before the improvement. The improved algorithm improves the accuracy of detection under the task of text detection in real scenes, showing the superiority of the detection algorithm.
在公共网络上广泛传播的视频不可避免地包含有隐私信息.为此很多针对视频的加密方案相继提出.文章针对HEVC视频感兴趣区域(ROI)提出一种自同步混沌流密码加密算法,以人脸所在区域作为感兴趣区域为例,实现对HEVC视频隐私区域的保护.首先,根据动力系统反控制原理,构建出无简并三维离散超混沌系统.在此基础上,提出一种自同步混沌流密码系统.然后,对HEVC视频ROI实现加密,具体如下.对于HEVC视频序列中的I帧,检测人脸所在区域并作为ROI;对于P帧,则利用视频流中压缩域信息追踪ROI.并将ROI信息转化为Tiles索引.利用自同步混沌流密码系统生成的伪随机序列对Tiles索引以及Tile内的语法元素MVD符号位、MVD值和QTC符号位进行加密.解码端的解密算法也同时详细给出.同时,借助补充增强信息(SEI)实现编解码端对ROI信息的同步.实验结果验证了所提方案的有效性.最后,对所提方法的性能和安全性进行了分析,并对所提方法的嵌入对编码器所带来的影响进行了讨论.
针对现有传统荧光材料防伪技术存在的可靠性较低、容易被仿制、防伪成本高昂等问题,文章提出了基于Faster R-CNN的荧光材料防伪图案识别方案,该方法使用了深度学习技术和区域提议网络(RPN)来实现对防伪图案的自动检测和识别.该方法首先使用区域提议网络生成候选图片区域,然后使用卷积神经网络对这些候选区域进行分类和定位,从而提高识别网络的分类准确率,缩减训练时间,图像识别准确率达 96%,获得了能对荧光材料防伪图案精准识别的网络模型,该方法在防伪图案识别方面具有很高的准确性和鲁棒性,可以有效地应用于荧光材料防伪图案识别领域,该研究为防伪技术的发展提供了新思路和方法.
由于现有的安全防护方法能耗量大,为此研究智能电网无线通信设备安全防护方法.根据不同的属性值生成决策树,计算信息的增益,得到最优的标签计算参数后获得其数组划分子集.对子集中的网络信息属性特征进行提取.选择最大增益值的数组,添加节点属性特征.进行攻击行为路径的预测,找到信息来源后识别安全威胁.对侧链区块数据进行预处理,将数据传递给到太坊主链上,并不断对其进行跟踪,数据异常时主链进行退回操作从而完成安全防护.实验结果表明,8 个小组的数据包在转发过程中,产生的能耗均小于 365MJ,使得通信数据在传输和转发中能够得到相对安全的空间,实现较好安全防护.
The construction of new electric power system is the key development direction of our national energy strategy,and its security is the top priority.Due to unstable energy supply,power load growth,meteorological disasters,man-made damage and other factors,the threats and challenges faced by the power system are increasingly complex.These challenges make tradi-tional approaches to power system security increasingly ineffective.Artificial intelligence technology is providing new ideas and solutions for power system security.To this end,this paper investigates the power system security application algorithms based on artificial intelligence technology,analyzes,compares and summarizes the power system security algorithms and their applications for three types of needs,namely,power system stability assessment,power security event recognition and process-ing,and power equipment detection and control.Then,the common technical challenges of current power system security algo-rithms are discussed.Finally,the development trend of power system security algorithm is predicted.
