Revisión de técnicas de aprendizaje de máquinas, usadas en el reconocimiento de especies vegetales
This research project describes the analysis of systematic literature review (RLS) with machine learning techniques in plant species. In RLS, an analysis of different articles was carried out, based on: recognition of plant species during the last 5 years and their frequency of use, for which search...
Збережено в:
| Автор: | |
|---|---|
| Інші автори: | |
| Формат: | bachelorThesis |
| Мова: | spa |
| Опубліковано: |
2022
|
| Предмети: | |
| Онлайн доступ: | http://repositorio.utc.edu.ec/handle/27000/9183 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
Додати тег | Резюме: | This research project describes the analysis of systematic literature review (RLS) with machine learning techniques in plant species. In RLS, an analysis of different articles was carried out, based on: recognition of plant species during the last 5 years and their frequency of use, for which search terms, selection of search protocols, and search resources were used in Science Direct, Academic Google and Scielo. The search selection was applied under quality criteria to the 23 articles of which 7 were approved, which will be used in the investigation. As a result, the mint and oregano plant species were chosen based on physical characteristics and, in turn, the logistic regression machine learning technique. For the development of the mobile application, a database of 500 photographs between mint and oregano was used, for the application of the logistic regression algorithm, 25 attributes resulting from the segmentation of images were used as input data (Convolution, Diffusion, Luminosity, Density, Threshold) using the filters Blur, Canny, Dilate, Eroded and the original image for each of them and two binary output data for determining the name of the plant. For the recognition of the plant, the function (y_i=β_0+β_1 x_i+ε_i for i=1,2,n(2,29) was used. Finally, the performance tests are carried out to verify the operation of the mobile application. |
|---|