QQ客服 官方微博 反馈留言

浏览/检索结果: 共5条，第1-5条 帮助 限定条件 专题：长春光学精密机械与物理研究所    第一署名单位    第一作者单位    通讯作者单位     已选(0)清除 条数/页：5101520253035404550556065707580859095100   排序方式：请选择作者升序作者降序题名升序题名降序发表日期升序发表日期降序提交时间升序提交时间降序 On hyperspectral remotely sensed image classification based on MNF and AdaBoosting (EI CONFERENCE) 会议论文 2012 3rd IEEE/IET International Conference on Audio, Language and Image Processing, ICALIP 2012, July 16, 2012 - July 18, 2012, Shanghai, China Xu Y.; Yu P.; Guo B.; Gao X.; Guo Y. 收藏  |  浏览/下载：16/0  |  提交时间：2013/03/25 As an effective statistical learning tool  AdaBoosting has been widely used in the field of pattern recognition. In this paper  a new method is proposed to improve the classification performance of hyperspectral images by combining the minimum noise fraction (MNF) and AdaBoosting. Because the hyperspectral imagery has many bands which have strong correlation and high redundancy  the hyperspectral data are pre-processed by the minimum noise fraction to reduce the data's dimensionality  whilst to remove noise bands simultaneously. Then  we use an AdaBoost algorithm to conduct the classification of hyperspectral remotely sensed image. Experimental results show that the classification accuracy is improved and the time of calculation is reduced as well. 2012 IEEE.   An improved hyperspectral classification algorithm based on back-propagation neural networks (EI CONFERENCE) 会议论文 2012 2nd International Conference on Remote Sensing, Environment and Transportation Engineering, RSETE 2012, June 1, 2012 - June 3, 2012, Nanjing, China Mao W.; Yu P.; Guo B.; Xu Y.; Chen H. 收藏  |  浏览/下载：19/0  |  提交时间：2013/03/25 In this paper  a new method is proposed to improve the classification performance of hyperspectral images by combining the principal component analysis (PCA)  genetic algorithm (GA)  and artificial neural networks (ANNs). First  some characteristics of the hyperspectral remotely sensed data  such as high correlation  high redundancy  etc.  are investigated. Based on the above analysis  we propose to use the principal component analysis to capture the main information existing in the hyperspectral images and reduce its dimensionality consequently. Next  we use neural networks to classify the reduced hyperspectral data. Since the back-propagation neural network we used is easy to suffer from the local minimum problem  we adopt a genetic algorithm to optimize the BP network's weights and the threshold. Experimental results show that the classification accuracy is improved and the time of calculation is reduced as well. 2012 IEEE.   Using bidirectional binary particle swarm optimization for feature selection in feature-level fusion recognition system (EI CONFERENCE) 会议论文 2009 4th IEEE Conference on Industrial Electronics and Applications, ICIEA 2009, May 25, 2009 - May 27, 2009, Xi'an, China Wang D.; Ge W.; Wang Y. 收藏  |  浏览/下载：16/0  |  提交时间：2013/03/25 In feature-level fusion recognition system  the other is optimizing system sensor design to get outstanding cost performance. So feature selection become usually necessary to reduce dimensionality of the combination of multi-sensor features and improve system performance in system design. In general  there are two main missions. One is improving the recognition correct rate as soon as possible  the optimization is usually applied to feature selection because of its computational feasibility and validity. For further improving recognition accuracy and reducing selected feature dimensions  this paper presents a more rational and accurate optimization  Bidirectional Binary Particle Swarm Optimization (BBPSO) algorithm for feature selection in feature-level fusion target recognition system. In addition  we introduce a new evaluating function as criterion function in BBPSO feature selection method. At the last  we utilized Leave-One-Out method to validate the proposed method. The experiment results show that the proposed algorithm improves classification accuracy by two percentage points  while the selected feature dimensions are less one dimension than original Particle Swarm Optimization approach with 16 original feature dimensions. 2009 IEEE.   Infrared face recognition using linear subspace analysis (EI CONFERENCE) 会议论文 MIPPR 2009 - Pattern Recognition and Computer Vision: 6th International Symposium on Multispectral Image Processing and Pattern Recognition, October 30, 2009 - November 1, 2009, Yichang, China Ge W.; Wang D.; Cheng Y.; Zhu M. 收藏  |  浏览/下载：26/0  |  提交时间：2013/03/25 Infrared image offers the main advantage over visible image of being invariant to illumination changes for face recognition. In this paper  based on the introduction of main methods of linear subspace analysis  such as Principal Component Analysis (PCA)  Linear Discriminant Analysis(LDA) and Fast Independent Component Analysis (FastICA)  the application of these methods to the recognition of infrared face images offered by OTCBVS workshop are investigated  and the advantages and disadvantages are compared. Experimental results show that the combination approach of PCA and LDA leads to better classification performance than single PCA approach or LDA approach  while the FastICA approach leads to the best classification performance with the improvement of nearly 5% compared with the combination approach. 2009 Copyright SPIE - The International Society for Optical Engineering.   Intelligent MRTD testing for thermal imaging system using ANN (EI CONFERENCE) 会议论文 ICO20: Remote Sensing and Infrared Devices and Systems, August 21, 2005 - August 26, 2005, Changchun, China Sun J.; Ma D. 收藏  |  浏览/下载：18/0  |  提交时间：2013/03/25 The Minimum Resolvable Temperature Difference (MRTD) is the most widely accepted figure for describing the performance of a thermal imaging system. Many models have been proposed to predict it. The MRTD testing is a psychophysical task  for which biases are unavoidable. It requires laboratory conditions such as normal air condition and a constant temperature. It also needs expensive measuring equipments and takes a considerable period of time. Especially when measuring imagers of the same type  the test is time consuming. So an automated and intelligent measurement method should be discussed. This paper adopts the concept of automated MRTD testing using boundary contour system and fuzzy ARTMAP  but uses different methods. It describes an Automated MRTD Testing procedure basing on Back-Propagation Network. Firstly  we use frame grabber to capture the 4-bar target image data. Then according to image gray scale  we segment the image to get 4-bar place and extract feature vector representing the image characteristic and human detection ability. These feature sets  along with known target visibility  are used to train the ANN (Artificial Neural Networks). Actually it is a nonlinear classification (of input dimensions) of the image series using ANN. Our task is to justify if image is resolvable or uncertainty. Then the trained ANN will emulate observer performance in determining MRTD. This method can reduce the uncertainties between observers and long time dependent factors by standardization. This paper will introduce the feature extraction algorithm  demonstrate the feasibility of the whole process and give the accuracy of MRTD measurement.