<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>In order to better extract scene and object information from computer image, a construction object extraction algorithm based on Bayesian network is proposed. The algorithm is trained by multi-scene aerial images to build a grain dictionary and map the grain in the actual image to the grain dictionary to obtain the scene information of the image;Then naive Bayesian networks were used to model the constraints of the relationship between architectural targets and the spatial context of scene classes, and the extraction of architectural targets was converted into a posteriori probability problem for solving Bayesian network class nodes. The experimental results show that the proposed algorithm can effectively extract architectural objects from aerial images. The experiment result shows that:In this paper, the proportion of target pixels accurately extracted by the algorithm is taken as the standard to define the standard of target pixels accurately extracted by the algorithm to reach more than 90% of the building target pixels. The average time of training an image is 2 s, which is mainly spent on the convolution operation with the filter. After the training, the average time of processing a single test image is 0.5s. It is proved that Bayesian network model can effectively extract scene and object information from computer image.</p>
</abstract>

In order to better extract scene and object information from computer image, a construction object extraction algorithm based on Bayesian network is proposed. The algorithm is trained by multi-scene aerial images to build a grain dictionary and map the grain in the actual image to the grain dictionary to obtain the scene information of the image;Then naive Bayesian networks were used to model the constraints of the relationship between architectural targets and the spatial context of scene classes, and the extraction of architectural targets was converted into a posteriori probability problem for solving Bayesian network class nodes. The experimental results show that the proposed algorithm can effectively extract architectural objects from aerial images. The experiment result shows that:In this paper, the proportion of target pixels accurately extracted by the algorithm is taken as the standard to define the standard of target pixels accurately extracted by the algorithm to reach more than 90% of the building target pixels. The average time of training an image is 2 s, which is mainly spent on the convolution operation with the filter. After the training, the average time of processing a single test image is 0.5s. It is proved that Bayesian network model can effectively extract scene and object information from computer image.

Computer Image Scene and Object Information Extraction based on Bayesian Network Model

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

In order to better extract scene and object information from computer image, a construction object extraction algorithm based on Bayesian network is proposed....

Computer Image Scene and Object Information Extraction based on Bayesian Network Model

Online veröffentlicht: 06. Juni 2023

Seitenbereich: 1859 - 1868

Eingereicht: 10. Juni 2022

Akzeptiert: 22. Nov. 2022

DOI: https://doi.org/10.2478/amns.2023.1.00289

Schlüsselwörter
Bayesian network model, Image scene, Extraction of information

© 2023 Hui Zhao et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Computer Image Scene and Object Information Extraction based on Bayesian Network Model

Online veröffentlicht: 06. Juni 2023

Seitenbereich: 1859 - 1868

Eingereicht: 10. Juni 2022

Akzeptiert: 22. Nov. 2022

DOI: https://doi.org/10.2478/amns.2023.1.00289

SchlüsselwörterBayesian network model, Image scene, Extraction of information

© 2023 Hui Zhao et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Schlüsselwörter
Bayesian network model, Image scene, Extraction of information