基於深度學習與圖像分割於衛星影像之農業經濟發展預測與視覺化

Economic Forecasting and Visualization in Agriculture via Semantic Segmentation of Satellite Images using Deep Learning

林宜安(Yi-An LIN)

指導教授 ： 高君豪

淡江大學/商管學院/大數據分析與商業智慧碩士學位學程/碩士(2022年)

https://doi.org/10.6846/TKU.2022.00679

摘要

從早期的台灣，在經過時間的演變裡，種植的設備和機器都有顯著的升級，近幾年在農產品的推廣下，使農產品運銷的效率增加，也讓台灣走向國際化的一面。而現今網際網路的發達，許多地圖軟體帶給人們生活上很大的方便性。本研究利用 Python 網路爬蟲，抓取地理資訊系統 Google Map 上的衛星空拍畫面。並以台灣農田部分做為研究的重點，使用深度學習預測模型和影像相關處理，透過建立模型預測在衛星圖片中所包含的農業面積，同時計算其經濟產值。根據預測結果，可節省政府於農地調查時所花費之人力，並且協助欲踏入農業開發事業之使用者，能夠對其土地之產值有概略了解並評估自身狀況。

關鍵字

深度學習 ； 圖像分類 ； 影像分割 ； 資料擴充 ； 預測模型

並列摘要

From the early days of Taiwan, over time, the planting equipment and machines have been significantly upgraded. In recent years, the promotion of agricultural products has increased the efficiency of agricultural product transportation and sales, and Taiwan has also become more international. With the development of the Internet nowadays,many map softwares bring great convenience to people's life. In this study,Python web crawler was used to capture satellite aerial images on the geographic information system Google Map. The research focuses on the farmland in Taiwan, using a deep learning prediction model and image-related processing to predict the agricultural area included in the satellite image by building a model,and calculate its economic output value at the same time. According to the forecast results, it can save the government's manpower in agricultural land surveys, and assist users who want to enter the agricultural development business to have a general understanding of the output value of their land and evaluate their own conditions.

並列關鍵字

Deep Learning ； Image Classification ； Image segmentation ； Data Augmentation ； Predictive Modelling

參考文獻

1. PK, F. A. (1984). What is Artificial Intelligence?. “Success is no accident. It is hard work, perseverance, learning, studying, sacrifice and most of all, love of what you are doing or learning to do”., 65.

2. Shetty, A., Thorat, A., Singru, R., Shigawan, M., Gaikwad, V. (2020, July). Predict Socio-Economic Status of an Area from Satellite Image Using Deep Learning. In 2020 International Conference on Electronics and Sustainable Communication Systems (ICESC) (pp. 177-182). IEEE.

3. Rout, A. R., Bagal, S. B. (2017, August). Natural Scene Classification Using Deep Learning. In 2017 International Conference on Computing, Communication, Control and Automation (ICCUBEA) (pp. 1-5). IEEE.

4. Liang, J., Deng, Y., Zeng, D. (2020). A deep neural network combined CNN and GCN for remote sensing scene classification. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 13, 4325-4338.

5. Shankar, T., Yamuna, G., Suman, G. (2013, April). Segmentation of natural colour image based on colour-texture features. In 2013 International Conference on Communication and Signal Processing (pp. 455-459). IEEE.