Use this link to cite:
http://hdl.handle.net/2183/31205 Aplicación de técnicas de intelixencia artificial para a realización automatizada de medidas en imaxes de produción madereira
Loading...
Identifiers
Publication date
Authors
Rojo Pose, Ronny
Advisors
Other responsabilities
Enxeñaría informática, Grao en
Journal Title
Bibliographic citation
Type of academic work
Academic degree
Abstract
[Resumo] Na produción madereira almacénanse grandes cargas de madeira formadas por táboas de distintos
tamaños, polo que a hora de traballar con este material, é necesario calcular a cantidade
real de madeira da forma máis precisa posible.
Este proceso, actualmente estase a levar a cabo manualmente, sen ter en conta as imprecisións
que se xerán durante o cálculo debido aos espazos que se poden crear entre as táboas
apiladas.
A automatización deste proceso, traballando directamente coas imaxes das propias cargas,
facilitará o traballo dos empleados. Esto permitirá acelerar parte do seu traballo, que é en
certo modo rutinario, permitindo mellorar tanto a velocidade do cálculo como a precisión
deste, para así poder fixar un prezo máis xusto acorde á carga de madeira.
Neste proxecto, enfocado no ámbito da Visión Artificial, investigarase o uso de diferentes
técnicas de procesado de imaxe para proporcionar posibles solucións ao problema descrito.
Entre estas propostas, empregaranse técnicas baseadas en filtros e técnicas de Intelixencia
Artificial, concretamente as redes neuronais artificiais. Ademais, desenvolverase unha aplicación
que presente unha interface gráfica, para permitir ao usuario utilizar o sistema creado.
[Abstract] In timber production, large loads of wood are stored. formed by boards of different sizes, so when working with this material, it is necessary to calculate the real amount of wood as accurately as possible. This process is currently carried out manually, without taking into account the inaccuracies that are generated during the calculation due to the spaces that can be created between the stacked boards. The automation of this process, working directly with the images of the loads themselves, will facilitate the work of the employees. This will speed up part of their work, which is somewhat routine, allowing them to improve both the speed of the calculation and its accuracy, so that they can set a fairer price according to the load of timber. In this project, focused on the field of Artificial Vision, the use of different image processing techniques will be investigated to provide possible solutions to the problem described. Among these proposals, we will use techniques based on filters and Artificial Intelligence techniques, specifically artificial neural networks. In addition, an application with a graphical interface will be developed to allow the user to use the created system.
[Abstract] In timber production, large loads of wood are stored. formed by boards of different sizes, so when working with this material, it is necessary to calculate the real amount of wood as accurately as possible. This process is currently carried out manually, without taking into account the inaccuracies that are generated during the calculation due to the spaces that can be created between the stacked boards. The automation of this process, working directly with the images of the loads themselves, will facilitate the work of the employees. This will speed up part of their work, which is somewhat routine, allowing them to improve both the speed of the calculation and its accuracy, so that they can set a fairer price according to the load of timber. In this project, focused on the field of Artificial Vision, the use of different image processing techniques will be investigated to provide possible solutions to the problem described. Among these proposals, we will use techniques based on filters and Artificial Intelligence techniques, specifically artificial neural networks. In addition, an application with a graphical interface will be developed to allow the user to use the created system.
Description
Editor version
Rights
Atribución-NoComercial-SinDerivadas 3.0 España








