AI SearchDeveloping a neural network based contextual classifier for cloud detection and making cloud free mosaic images
Call Number: AIT Thesis no. CS-91-30 Material type:
TextSeries: Asian Institute of Technology. Thesis ; no. CS-91-30Publication details: Bangkok : Asian Institute of Technology, 1991Description: 56 leavesSubject(s): Online resources: Dissertation note: Thesis (M.Eng.) - Asian Institute of Technology, 1991 Summary: Certain areas of the earth's surface are constantly covered by clouds during most of the time of the year. Obtaining cloud free images of such areas is an extremely difficult task. Neural networks can be meaningfully used as classifiers in situations where the data to be classified is of non parametric nature. This will be the situation encountered when we consider all the clouds as one class and all. the clear sky areas as another class. The classification accuracy can be increased if we take in to account the contextual dependencies between objects present in the image. The theory of statistical contextual classification is well developed and has been already successfully applied to cloud detection and classification. In this study we extend the contextual classification algorithms to neural networks. This method relives us from calculating complex class conditional distributions, and does not rely on the parametric nature of the distributions. The contextual features have been used to reduce the instances of misclassification produced by a pixel by pixel neural network cloud detector. This experimental part of this study has been carried out with 50m resolution MOS-1 (Marine Observation Satellite) data. A relatively cloud free mosaic image has also been made.
| Cover image | Item type | Current library | Home library | Collection | Shelving location | Call number | Materials specified | Vol info | URL | Copy number | Status | Notes | Date due | Barcode | Item holds | Item hold queue priority | Course reserves | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
22-AIT Thesis (Replacement)
|
Asian Institute of Technology Library AIT Publications | AIT Thesis no. CS-91-30 (Browse shelf(Opens below)) | 3 | Available | 30050120676894 | |||||||||||||
40-Archives
|
Asian Institute of Technology Library Archives | AIT Thesis no. CS-91-30 (Browse shelf(Opens below)) | Available | 30050120353320 |
A thesis submitted in partial fulfillment of the requirement for the degree of master of engineering, School of Engineering and Technology
Thesis (M.Eng.) - Asian Institute of Technology, 1991
Certain areas of the earth's surface are constantly covered by clouds during most of the time of the year. Obtaining cloud free images of such areas is an extremely difficult task. Neural networks can be meaningfully used as classifiers in situations where the data to be classified is of non parametric nature. This will be the situation encountered when we consider all the clouds as one class and all. the clear sky areas as another class. The classification accuracy can be increased if we take in to account the contextual dependencies between objects present in the image. The theory of statistical contextual classification is well developed and has been already successfully applied to cloud detection and classification. In this study we extend the contextual classification algorithms to neural networks. This method relives us from calculating complex class conditional distributions, and does not rely on the parametric nature of the distributions. The contextual features have been used to reduce the instances of misclassification produced by a pixel by pixel neural network cloud detector. This experimental part of this study has been carried out with 50m resolution MOS-1 (Marine Observation Satellite) data. A relatively cloud free mosaic image has also been made.
There are no comments on this title.