Pub Date : 2022-11-01DOI: 10.1109/CCISP55629.2022.9974309
Xiaohui Jin, Fuyou Miao
In a $(k, n)$ threshold secret image sharing (SIS) scheme, the original secret image can be reconstructed losslessly by collecting no less than $k$ shadow images without errors. However, during the transmission or storage of the shadow image, if the transmission channel is noisy or the storage medium is unreliable, the shadow images will be erroneous. In this case, most of the existing SIS schemes cannot reconstruct the secret image. In an effort to cope with the issue, we propose a robust SIS (RSIS) scheme based on the Robust Chinese Remainder Theorem (RCRT), namely RSIS-RCRT, which can realize the reconstruction of the secret image with high quality under a certain error in the shadow image. In addition, a simple and effective Image Correction Method (ICM) is proposed, which improves the quality of the reconstructed image significantly in RSIS-RCRT scheme. Experiments show that the RSIS-RCRT scheme and ICM are effective.
在$(k， n)$阈值秘密图像共享(SIS)方案中，通过收集不少于$k$的阴影图像，可以无损地重建原始秘密图像。然而，在阴影图像的传输或存储过程中，如果传输信道有噪声或存储介质不可靠，则会产生错误的阴影图像。在这种情况下，大多数现有的SIS方案都无法重建秘密图像。为了解决这一问题，我们提出了一种基于鲁棒中国剩余定理(robust Chinese residual Theorem, RCRT)的鲁棒SIS (RSIS)方案，即RSIS-RCRT，可以在阴影图像一定误差下实现高质量的秘密图像重建。此外，提出了一种简单有效的图像校正方法(ICM)，显著提高了rss - rcrt方案重建图像的质量。实验表明，RSIS-RCRT方案和ICM方案是有效的。
{"title":"Robust Secret Image Sharing Based on Robust Chinese Reminder Theorem","authors":"Xiaohui Jin, Fuyou Miao","doi":"10.1109/CCISP55629.2022.9974309","DOIUrl":"https://doi.org/10.1109/CCISP55629.2022.9974309","url":null,"abstract":"In a $(k, n)$ threshold secret image sharing (SIS) scheme, the original secret image can be reconstructed losslessly by collecting no less than $k$ shadow images without errors. However, during the transmission or storage of the shadow image, if the transmission channel is noisy or the storage medium is unreliable, the shadow images will be erroneous. In this case, most of the existing SIS schemes cannot reconstruct the secret image. In an effort to cope with the issue, we propose a robust SIS (RSIS) scheme based on the Robust Chinese Remainder Theorem (RCRT), namely RSIS-RCRT, which can realize the reconstruction of the secret image with high quality under a certain error in the shadow image. In addition, a simple and effective Image Correction Method (ICM) is proposed, which improves the quality of the reconstructed image significantly in RSIS-RCRT scheme. Experiments show that the RSIS-RCRT scheme and ICM are effective.","PeriodicalId":431851,"journal":{"name":"2022 7th International Conference on Communication, Image and Signal Processing (CCISP)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134535231","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-01DOI: 10.1109/CCISP55629.2022.9974259
Xianli Zhou, Wenwen Zhu, F. Liu, Wei Yang, Miao Chu
As the target moves in the background, the profile dislocation will be caused and the camouflage effect will be reduced following that. Aimed at the problem, the author in the research introduces a camouflage evaluation method that based on the similarity of the edge contour of the camouflage target. The research combines Sobel and Canny algorithm to extract the edge contour of the target, uses the Euclidean distance to calculate the similarity of target's edge contour to its background after camouflage, so as to judge the quality of the camouflage effect. Finally, the rationality of the method is verified by ranking and comparing six camouflage plans with different contours, which were calculated by edge contour similarity method and image discrimination method respectively.
{"title":"The Evaluation of Camouflage Based on Image Edge Contour Similarity","authors":"Xianli Zhou, Wenwen Zhu, F. Liu, Wei Yang, Miao Chu","doi":"10.1109/CCISP55629.2022.9974259","DOIUrl":"https://doi.org/10.1109/CCISP55629.2022.9974259","url":null,"abstract":"As the target moves in the background, the profile dislocation will be caused and the camouflage effect will be reduced following that. Aimed at the problem, the author in the research introduces a camouflage evaluation method that based on the similarity of the edge contour of the camouflage target. The research combines Sobel and Canny algorithm to extract the edge contour of the target, uses the Euclidean distance to calculate the similarity of target's edge contour to its background after camouflage, so as to judge the quality of the camouflage effect. Finally, the rationality of the method is verified by ranking and comparing six camouflage plans with different contours, which were calculated by edge contour similarity method and image discrimination method respectively.","PeriodicalId":431851,"journal":{"name":"2022 7th International Conference on Communication, Image and Signal Processing (CCISP)","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133775083","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-01DOI: 10.1109/CCISP55629.2022.9974340
Zhao Feng
As China gradually enters the “5G+artificial” intelligent information era, the vigorous development of emerging technologies such as artificial intelligence, face recognition technology has been widely used to education, security and finance and other fields, at the same time, the collection of personal face recognition information is quietly leaked, caused concern and anxiety from public. In the special group of colleges and universities, the legitimacy of collecting face recognition information is controversial. This paper centers on the different expressions of collecting face recognition information in colleges and universities.
{"title":"The Problem of Collecting Face Recognition Information in Colleges and Universities","authors":"Zhao Feng","doi":"10.1109/CCISP55629.2022.9974340","DOIUrl":"https://doi.org/10.1109/CCISP55629.2022.9974340","url":null,"abstract":"As China gradually enters the “5G+artificial” intelligent information era, the vigorous development of emerging technologies such as artificial intelligence, face recognition technology has been widely used to education, security and finance and other fields, at the same time, the collection of personal face recognition information is quietly leaked, caused concern and anxiety from public. In the special group of colleges and universities, the legitimacy of collecting face recognition information is controversial. This paper centers on the different expressions of collecting face recognition information in colleges and universities.","PeriodicalId":431851,"journal":{"name":"2022 7th International Conference on Communication, Image and Signal Processing (CCISP)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123799823","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Multi-station directional crossover technique is a commonly used passive localization technique. In contrast to this technique, in the field of radio localization of sensor network nodes, multiple radiation sources with known locations can also be used to localize nodes with unknown locations that are capable of receiving radio signals emitted from the sources, i.e., the nodal direction finding cross-location method. The nodal direction finding cross-location method requires attitude measurement, and since attitude measurement has errors and increases the size, weight and cost of the node. This paper proposes to use the incoming azimuth difference measured at the node to localize the node, which does not require attitude measurement. We verify through simulation that the method ensures localization accuracy while saving attitude measurement, which is of great importance in engineering practice. The influence of various measurement parameters in the positioning system on positioning accuracy is also analyzed, among which orientation accuracy is the main influencing factor of positioning accuracy.
{"title":"Passive Positioning from Azimuth Difference of Opportunistic Signals Without Attitude Measurement","authors":"Junlin Li, Maoyan Ran, Ying He, Jinke Yang, Xiongfei Li, Q. Wan","doi":"10.1109/CCISP55629.2022.9974331","DOIUrl":"https://doi.org/10.1109/CCISP55629.2022.9974331","url":null,"abstract":"Multi-station directional crossover technique is a commonly used passive localization technique. In contrast to this technique, in the field of radio localization of sensor network nodes, multiple radiation sources with known locations can also be used to localize nodes with unknown locations that are capable of receiving radio signals emitted from the sources, i.e., the nodal direction finding cross-location method. The nodal direction finding cross-location method requires attitude measurement, and since attitude measurement has errors and increases the size, weight and cost of the node. This paper proposes to use the incoming azimuth difference measured at the node to localize the node, which does not require attitude measurement. We verify through simulation that the method ensures localization accuracy while saving attitude measurement, which is of great importance in engineering practice. The influence of various measurement parameters in the positioning system on positioning accuracy is also analyzed, among which orientation accuracy is the main influencing factor of positioning accuracy.","PeriodicalId":431851,"journal":{"name":"2022 7th International Conference on Communication, Image and Signal Processing (CCISP)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125108413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-01DOI: 10.1109/CCISP55629.2022.9974546
Xinyi Sun
Food is a part of life, and it is something that everyone needs to consider. Everyone's time is precious in this busy modern city, and canteen efficiency has always been a criticism, with long lines taking up most of the time. The fully automatic dish recognition system can effectively solve this problem and reduce labor costs for the company. In this paper, the research work focuses on solving the problem of recognizing dishes in a fully automated dish recognition system and mainly employs a deep learning method of convolutional neural networks to identify different kinds of dishes. Firstly, the original images are pre-processed and divided into test and train data sets. Then the tensor flow is used to build a network model based on a convolutional neural network. The recognition accuracy reaches 81.21% after the algorithm's overall optimization and the parameters' adjustment in several trials.
{"title":"The Application of Neural Network in Dish Recognition","authors":"Xinyi Sun","doi":"10.1109/CCISP55629.2022.9974546","DOIUrl":"https://doi.org/10.1109/CCISP55629.2022.9974546","url":null,"abstract":"Food is a part of life, and it is something that everyone needs to consider. Everyone's time is precious in this busy modern city, and canteen efficiency has always been a criticism, with long lines taking up most of the time. The fully automatic dish recognition system can effectively solve this problem and reduce labor costs for the company. In this paper, the research work focuses on solving the problem of recognizing dishes in a fully automated dish recognition system and mainly employs a deep learning method of convolutional neural networks to identify different kinds of dishes. Firstly, the original images are pre-processed and divided into test and train data sets. Then the tensor flow is used to build a network model based on a convolutional neural network. The recognition accuracy reaches 81.21% after the algorithm's overall optimization and the parameters' adjustment in several trials.","PeriodicalId":431851,"journal":{"name":"2022 7th International Conference on Communication, Image and Signal Processing (CCISP)","volume":"86 6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133834026","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-01DOI: 10.1109/CCISP55629.2022.9974584
Qiansheng Fang, Qiyu Li, Liangliang Su, Yalong Yang
Sketches are usually composed of simple strokes. Compared with the natural image, they lack the information of texture and color. Most of the existing works do not reduce the impact of blank region in sketches very well. This paper proposes a new deep convolutional neural network named Sketch Fusion Net (SFN) model to minimize the effect of the blank region and focus on information region. This model is mainly composed of Sketch Block Convolutional (SBConv) modules, the SBCnov module integrates an attention mechanism and an improved residual network. The first one can effectively extract the effective part of the sketch information. The other one uses multi-level feature combination strategy to extract richer semantic information, which can effectively relieve the problem of model degradation. Finally, two public datasets, namely the TU-Berlin and Sketchy, are used for sketch recognition experiments. The results demonstrate that this proposed method improves the recognition accuracy by 2.1% and 7.2% over several state-of-the-art methods and yields promising results.
{"title":"Sketch recognition based on attention mechanism and improved residual network","authors":"Qiansheng Fang, Qiyu Li, Liangliang Su, Yalong Yang","doi":"10.1109/CCISP55629.2022.9974584","DOIUrl":"https://doi.org/10.1109/CCISP55629.2022.9974584","url":null,"abstract":"Sketches are usually composed of simple strokes. Compared with the natural image, they lack the information of texture and color. Most of the existing works do not reduce the impact of blank region in sketches very well. This paper proposes a new deep convolutional neural network named Sketch Fusion Net (SFN) model to minimize the effect of the blank region and focus on information region. This model is mainly composed of Sketch Block Convolutional (SBConv) modules, the SBCnov module integrates an attention mechanism and an improved residual network. The first one can effectively extract the effective part of the sketch information. The other one uses multi-level feature combination strategy to extract richer semantic information, which can effectively relieve the problem of model degradation. Finally, two public datasets, namely the TU-Berlin and Sketchy, are used for sketch recognition experiments. The results demonstrate that this proposed method improves the recognition accuracy by 2.1% and 7.2% over several state-of-the-art methods and yields promising results.","PeriodicalId":431851,"journal":{"name":"2022 7th International Conference on Communication, Image and Signal Processing (CCISP)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121962634","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-01DOI: 10.1109/CCISP55629.2022.9974205
Yue Wang, Aijun Wang, Honglei Xu, W. Wang, Huanan Li, Zhen-fu Cui
With the development of SDN technology, PCEP has become a path computing protocol widely used in the existing network. It takes the controller as the core of the overall architecture, and can be applied to calculate constrained paths in cross layer and cross domain environments in complex networks. According to the discussion on the application of PCEP in cloud-network integration scenarios, this paper proposes a VLAN-based traffic assurance mechanism based on the extension of PCEP. It can meet the requirements of key service assurance in the native IP environment and establish a connection oriented network tunnel. Under the VLAN-based architecture based on PCEP, the operator can perform closed-loop automatic control of the network and the mechanism helps to improve the intelligent scheduling of the network and the ability of real-time perception which meets the maintenance and operation requirements like flexible architecture, comprehensive opening of capabilities and global scheduling of resources.
{"title":"Research on PCEP Extension for VLAN-based Traffic Forwarding in cloud network integration","authors":"Yue Wang, Aijun Wang, Honglei Xu, W. Wang, Huanan Li, Zhen-fu Cui","doi":"10.1109/CCISP55629.2022.9974205","DOIUrl":"https://doi.org/10.1109/CCISP55629.2022.9974205","url":null,"abstract":"With the development of SDN technology, PCEP has become a path computing protocol widely used in the existing network. It takes the controller as the core of the overall architecture, and can be applied to calculate constrained paths in cross layer and cross domain environments in complex networks. According to the discussion on the application of PCEP in cloud-network integration scenarios, this paper proposes a VLAN-based traffic assurance mechanism based on the extension of PCEP. It can meet the requirements of key service assurance in the native IP environment and establish a connection oriented network tunnel. Under the VLAN-based architecture based on PCEP, the operator can perform closed-loop automatic control of the network and the mechanism helps to improve the intelligent scheduling of the network and the ability of real-time perception which meets the maintenance and operation requirements like flexible architecture, comprehensive opening of capabilities and global scheduling of resources.","PeriodicalId":431851,"journal":{"name":"2022 7th International Conference on Communication, Image and Signal Processing (CCISP)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114078866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-01DOI: 10.1109/CCISP55629.2022.9974294
Wei Lv, Na Sun, Wenxi Ni, Shen Shen
In order to research the underwater bubble screen echo, a modeling and simulation method is proposed in this paper. First, the mathematical model of underwater bubble shell is modeled. Second, the simulation experiments are carried out in different working conditions. Finally, the presented model is validated by comparing the spatial-temporal characteristics of the simulation data and the trial data using the space-time adaptive process.
{"title":"Simulation and Spatial-Temporal Characteristics Analysis of Underwater Bubble Screen Echo","authors":"Wei Lv, Na Sun, Wenxi Ni, Shen Shen","doi":"10.1109/CCISP55629.2022.9974294","DOIUrl":"https://doi.org/10.1109/CCISP55629.2022.9974294","url":null,"abstract":"In order to research the underwater bubble screen echo, a modeling and simulation method is proposed in this paper. First, the mathematical model of underwater bubble shell is modeled. Second, the simulation experiments are carried out in different working conditions. Finally, the presented model is validated by comparing the spatial-temporal characteristics of the simulation data and the trial data using the space-time adaptive process.","PeriodicalId":431851,"journal":{"name":"2022 7th International Conference on Communication, Image and Signal Processing (CCISP)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131931924","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-01DOI: 10.1109/CCISP55629.2022.9974291
C. Du, Chao Zhang, Xiang Peng, Hong Guo
Magnetic anomaly detection (MAD) is a method to find the ferromagnetic object by recognizing the weak target magnetic signal in the complex background magnetic noise. In the practical detection, background magnetic noise is usually complex colored noise, so the magnetic noise needs firstly to be suppressed before detecting the measurement magnetic data. In this paper, singular spectrum analysis (SSA) method is firstly introduced to decompose the test data to improve the signal-to-noise ratio (SNR) of detection data. In the decomposing procedure, the clustering method is used to classify the singular values and select out the singular values containing the information of target to reconstruct the new signal to be detected. And then the orthogonal basis functions (OBFs) is applied to detect the reconstructed signal considering that the OBFs has strong resistance to white noise. Some simulation experiments were conducted to show that the detection probability of this method in this paper for the target signal submerged in colored noise is improved by more than 25% compared with the traditional OBFs detection algorithm.
{"title":"Magnetic Anomaly Detection Based on Singular Spectrum Analysis and Orthonormal Basis Functions","authors":"C. Du, Chao Zhang, Xiang Peng, Hong Guo","doi":"10.1109/CCISP55629.2022.9974291","DOIUrl":"https://doi.org/10.1109/CCISP55629.2022.9974291","url":null,"abstract":"Magnetic anomaly detection (MAD) is a method to find the ferromagnetic object by recognizing the weak target magnetic signal in the complex background magnetic noise. In the practical detection, background magnetic noise is usually complex colored noise, so the magnetic noise needs firstly to be suppressed before detecting the measurement magnetic data. In this paper, singular spectrum analysis (SSA) method is firstly introduced to decompose the test data to improve the signal-to-noise ratio (SNR) of detection data. In the decomposing procedure, the clustering method is used to classify the singular values and select out the singular values containing the information of target to reconstruct the new signal to be detected. And then the orthogonal basis functions (OBFs) is applied to detect the reconstructed signal considering that the OBFs has strong resistance to white noise. Some simulation experiments were conducted to show that the detection probability of this method in this paper for the target signal submerged in colored noise is improved by more than 25% compared with the traditional OBFs detection algorithm.","PeriodicalId":431851,"journal":{"name":"2022 7th International Conference on Communication, Image and Signal Processing (CCISP)","volume":"105 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121106990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The integration of infrared and visible images can take advantage of temperature information from infrared modality and sharp appearance from visible modality, and therefore it is helpful to improve the accuracy of localization and fault diagnosis of electrical equipment. A key step towards integration analysis is to register the images in infrared and visible modalities. In this paper, we propose a new method for infrared and visible image registration. In order to deal with large difference between these two modalities, we first transform both infrared and visible images into radiation-invariant maps. Then, LoFTR, which is a self-attention based deep neural network, is adopted to extract and match features based on the radiation-invariant maps. Finally, we utilize a progressive sample consensus (PROSAC) algorithm to estimate the transformation parameters, based on which the infrared image can be transformed into the corresponding visible image coordinates. Experiments on an electrical equipment dataset show that our proposed method is robust to both radiation and geometric variations.
{"title":"A Robust Method for Electrical Equipment Infrared and Visible Image Registration","authors":"Ying Lin, Fengda Zhang, Meng Liu, Zhuangzhuang Li, Wenjie Zheng, Yi Yamg","doi":"10.1109/CCISP55629.2022.9974532","DOIUrl":"https://doi.org/10.1109/CCISP55629.2022.9974532","url":null,"abstract":"The integration of infrared and visible images can take advantage of temperature information from infrared modality and sharp appearance from visible modality, and therefore it is helpful to improve the accuracy of localization and fault diagnosis of electrical equipment. A key step towards integration analysis is to register the images in infrared and visible modalities. In this paper, we propose a new method for infrared and visible image registration. In order to deal with large difference between these two modalities, we first transform both infrared and visible images into radiation-invariant maps. Then, LoFTR, which is a self-attention based deep neural network, is adopted to extract and match features based on the radiation-invariant maps. Finally, we utilize a progressive sample consensus (PROSAC) algorithm to estimate the transformation parameters, based on which the infrared image can be transformed into the corresponding visible image coordinates. Experiments on an electrical equipment dataset show that our proposed method is robust to both radiation and geometric variations.","PeriodicalId":431851,"journal":{"name":"2022 7th International Conference on Communication, Image and Signal Processing (CCISP)","volume":"362 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126700117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：