• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • Tagged with
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Resolu??o em tomografia de tempo de tr?nsito po?o-a-po?o: a depend?ncia da ilumina??o

Dantas, Renato Ramos da Silva 05 March 2015 (has links)
Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2016-07-08T21:26:34Z No. of bitstreams: 1 RenatoRamosDaSilvaDantas_DISSERT.pdf: 7844658 bytes, checksum: 9553a9302f1452c512e5cb94621a0e65 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2016-07-12T21:56:07Z (GMT) No. of bitstreams: 1 RenatoRamosDaSilvaDantas_DISSERT.pdf: 7844658 bytes, checksum: 9553a9302f1452c512e5cb94621a0e65 (MD5) / Made available in DSpace on 2016-07-12T21:56:07Z (GMT). No. of bitstreams: 1 RenatoRamosDaSilvaDantas_DISSERT.pdf: 7844658 bytes, checksum: 9553a9302f1452c512e5cb94621a0e65 (MD5) Previous issue date: 2015-03-05 / Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico (CNPq) / O aspecto fundamental na resolu??o limitante em tomografia de tr?nsito po?o-a-po?o ? a ilumina??o,um resultado bem conhecido mas n?o t?o bem exemplificado. A resolu??o no caso 2D? revisitada usando uma simples abordagem geom?trica baseada na distribui??o de aberturas angularese nas propriedades da Transformada de Radon. Analiticamente ? mostrado que se umainterface tem mergulhos contidos nos limites da abertura angular em todos os pontos, ela ? corretamenteimageada no tomograma. Por invers?o de dados sint?ticos, esse resultado ? confirmadoe ? tamb?m evidenciado que artefatos isolados podem estar presentes quando o mergulho estiverpr?ximo do limite de ilumina??o. No sentido inverso, entretanto, se uma interface ? interpret?velpor um tomograma, mesmo uma interface aproximadamente horizontal, n?o h? garantia de queela corresponda a uma interface verdadeira. De modo semelhante, se um corpo estiver presente naregi?o entre os po?os, ele ? imageado no tomograma de forma difusa, mas suas interfaces - em particular,as bordas verticais - podem n?o ser resolvidas, e artefatos adicionais podem estar presentes.Novamente, no sentido inverso, n?o h? garantia que uma anomalia isolada corresponda a um corpoan?malo verdadeiro, pois sua anomalia pode tamb?m ser um artefato. Juntos, esses resultadosdeclaram o dilema dos problemas inversos mal-postos: aus?ncia de garantia de correspond?ncia? distribui??o verdadeira. As limita??es devidas ? ilumina??o podem n?o ser resolvidas pelo usode v?nculos matem?ticos. ? mostrado que tomogramas po?o-a-po?o derivadas pelo uso de v?nculosde esparsidade, usando tanto a Transformada de Cosseno Discreto como as bases Daubechies,basicamente reproduzem as mesmas caracter?sticas vistas em tomogramas obtidos com o v?nculode suavidade cl?ssico. ? necess?rio que interpreta??es sejam feitas sempre levando em considera??oas informa??es a priori e as limita??es particulares devido ? ilumina??o. Um exemplo deinterpreta??o de um levantamento de dados reais dentro deste contexto tamb?m ? apresentado. / The key aspect limiting resolution in crosswell traveltime tomography is illumination, a well known result but not as well exemplified. Resolution in the 2D case is revisited using a simple geometric approach based on the angular aperture distribution and the Radon Transform properties. Analitically it is shown that if an interface has dips contained in the angular aperture limits in all points, it is correctly imaged in the tomogram. By inversion of synthetic data this result is confirmed and it is also evidenced that isolated artifacts might be present when the dip is near the illumination limit. In the inverse sense, however, if an interface is interpretable from a tomogram, even an aproximately horizontal interface, there is no guarantee that it corresponds to a true interface. Similarly, if a body is present in the interwell region it is diffusely imaged in the tomogram, but its interfaces - particularly vertical edges - can not be resolved and additional artifacts might be present. Again, in the inverse sense, there is no guarantee that an isolated anomaly corresponds to a true anomalous body because this anomaly can also be an artifact. Jointly, these results state the dilemma of ill-posed inverse problems: absence of guarantee of correspondence to the true distribution. The limitations due to illumination may not be solved by the use of mathematical constraints. It is shown that crosswell tomograms derived by the use of sparsity constraints, using both Discrete Cosine Transform and Daubechies bases, basically reproduces the same features seen in tomograms obtained with the classic smoothness constraint. Interpretation must be done always taking in consideration the a priori information and the particular limitations due to illumination. An example of interpreting a real data survey in this context is also presented.

Page generated in 0.0696 seconds