FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W166983968> ?p ?o ?g. }

• W166983968 abstract "Image fusion is the process of combining images of a scene obtained from multiple sensors to obtain a single composite image. The goal is to reliably integrate image information from multisensor images to aid in tasks such as navigation guidance, object detection and recognition, medical diagnosis and data compression. The main challenges in fusion are caused primarily by local contrast reversals, mismatched sensor-specific image features and noise present in multisensor images. One or more of these conditions adversely affect existing fusion techniques. In this thesis, we present a probabilistic model-based approach for the fusion of multisensor images that addresses the shortcomings of existing solutions. We formulate the fusion task as a problem of estimating an underlying true scene from the sensor images. We model the sensor images as noisy, locally affine functions of this true scene. The parameters of the affine functions explicitly incorporate reversal in local contrast and the presence of sensor-specific image features in the sensor images. Given this model, we use a Bayesian framework to provide either maximum likelihood or maximum a posteriori estimates of the true scene from the sensor images. The estimate of the true scene constitutes our probabilistic fusion rule which resembles principal component projections. The fused image obtained by this rule is a locally weighted linear combination of the sensor images. The weights depend upon the parameters of the affine functions and the noise. The weights scale the sensor images according to the signal and noise content. We derive estimates of the model parameters from the sensor images. The least squares estimates of the affine parameters are based on the local covariance of the image data, and are related to local principal components analysis. Our fusion approach also incorporates prior image information about the scene. The contribution of the prior image information is locally weighted and added to the combination of the sensor images. The weighting determines the confidence in the prior. The inclusion of the prior provides the ability to obtain reliable fusion results when the sensor images are unreliable. We demonstrate the efficacy of our fusion approach on real and simulated images from visible-band and infrared sensors. We compare the results and computational complexity with those of the existing fusion techniques which are based on selection and averaging strategies. The results presented in this thesis illustrate that our probabilistic approach yields results that are similar to existing techniques when the noise is low and performs better than existing techniques when the noise is high. Common features and contrast reversed features are preserved, and sensor-specific features from each sensor image are retained in the fused image. The results using prior image information demonstrate that inclusion of prior information produces more reliable fused images." @default.
• W166983968 created "2016-06-24" @default.
• W166983968 creator A5000714396 @default.
• W166983968 creator A5020862899 @default.
• W166983968 creator A5034783699 @default.
• W166983968 date "1999-01-01" @default.
• W166983968 modified "2023-10-10" @default.
• W166983968 title "Probabilistic model-based multisensor image fusion" @default.
• W166983968 cites W13740476 @default.
• W166983968 cites W1535715006 @default.
• W166983968 cites W1544057354 @default.
• W166983968 cites W1568281349 @default.
• W166983968 cites W1575210522 @default.
• W166983968 cites W1586131326 @default.
• W166983968 cites W1600895443 @default.
• W166983968 cites W1601835124 @default.
• W166983968 cites W1622427076 @default.
• W166983968 cites W1836454392 @default.
• W166983968 cites W1963623641 @default.
• W166983968 cites W1964257080 @default.
• W166983968 cites W1966382716 @default.
• W166983968 cites W1973667959 @default.
• W166983968 cites W1977793627 @default.
• W166983968 cites W1979059554 @default.
• W166983968 cites W1988302498 @default.
• W166983968 cites W2003803873 @default.
• W166983968 cites W2018106792 @default.
• W166983968 cites W2024750028 @default.
• W166983968 cites W2034354530 @default.
• W166983968 cites W2038359451 @default.
• W166983968 cites W204502587 @default.
• W166983968 cites W2047506309 @default.
• W166983968 cites W2048913467 @default.
• W166983968 cites W2050320786 @default.
• W166983968 cites W2058929230 @default.
• W166983968 cites W2061146081 @default.
• W166983968 cites W2067710092 @default.
• W166983968 cites W2086153413 @default.
• W166983968 cites W2087380704 @default.
• W166983968 cites W2087987522 @default.
• W166983968 cites W2103504761 @default.
• W166983968 cites W2116044718 @default.
• W166983968 cites W2125027820 @default.
• W166983968 cites W2130657708 @default.
• W166983968 cites W2132746638 @default.
• W166983968 cites W2135346934 @default.
• W166983968 cites W2138714160 @default.
• W166983968 cites W2141899530 @default.
• W166983968 cites W2146610201 @default.
• W166983968 cites W2155986238 @default.
• W166983968 cites W2168393971 @default.
• W166983968 cites W2171396536 @default.
• W166983968 cites W2296589133 @default.
• W166983968 cites W2911709767 @default.
• W166983968 cites W599033465 @default.
• W166983968 cites W601809726 @default.
• W166983968 cites W649289763 @default.
• W166983968 cites W98769269 @default.
• W166983968 cites W2028555274 @default.
• W166983968 cites W3124884148 @default.
• W166983968 hasPublicationYear "1999" @default.
• W166983968 type Work @default.
• W166983968 sameAs 166983968 @default.
• W166983968 citedByCount "20" @default.
• W166983968 countsByYear W1669839682012 @default.
• W166983968 countsByYear W1669839682013 @default.
• W166983968 countsByYear W1669839682014 @default.
• W166983968 countsByYear W1669839682016 @default.
• W166983968 crossrefType "journal-article" @default.
• W166983968 hasAuthorship W166983968A5000714396 @default.
• W166983968 hasAuthorship W166983968A5020862899 @default.
• W166983968 hasAuthorship W166983968A5034783699 @default.
• W166983968 hasConcept C105795698 @default.
• W166983968 hasConcept C114289077 @default.
• W166983968 hasConcept C115961682 @default.
• W166983968 hasConcept C153180895 @default.
• W166983968 hasConcept C154945302 @default.
• W166983968 hasConcept C202444582 @default.
• W166983968 hasConcept C31972630 @default.
• W166983968 hasConcept C33923547 @default.
• W166983968 hasConcept C33954974 @default.
• W166983968 hasConcept C41008148 @default.
• W166983968 hasConcept C49781872 @default.
• W166983968 hasConcept C49937458 @default.
• W166983968 hasConcept C69744172 @default.
• W166983968 hasConcept C92757383 @default.
• W166983968 hasConcept C9810830 @default.
• W166983968 hasConcept C99498987 @default.
• W166983968 hasConceptScore W166983968C105795698 @default.
• W166983968 hasConceptScore W166983968C114289077 @default.
• W166983968 hasConceptScore W166983968C115961682 @default.
• W166983968 hasConceptScore W166983968C153180895 @default.
• W166983968 hasConceptScore W166983968C154945302 @default.
• W166983968 hasConceptScore W166983968C202444582 @default.
• W166983968 hasConceptScore W166983968C31972630 @default.
• W166983968 hasConceptScore W166983968C33923547 @default.
• W166983968 hasConceptScore W166983968C33954974 @default.
• W166983968 hasConceptScore W166983968C41008148 @default.
• W166983968 hasConceptScore W166983968C49781872 @default.
• W166983968 hasConceptScore W166983968C49937458 @default.

• next