基於影像辨識技術之蕃茄熟成及作物異常偵測系統

學號：M10756004 論文名稱：基於影像辨識技術之蕃茄熟成及作物異常偵測系統總頁數：51 學校名稱：國立屏東科技大學　　　　　　系別：資訊管理系畢業時間及摘要別：109學年度第一學期碩士學位論文摘要研究生：張惟智　　　　　　　　　　指導教授：蔡玉娟　博士論文摘要內容：本研究是基於影像辨識技術為基礎設計並建置一套「以影像辨識為基礎之蕃茄熟成度與作物病蟲害偵測系統」，主要目的：(1) 辨識蕃茄熟成狀況，利用攝相機進行影像拍攝正常熟成與未熟成蕃茄果實之影像。(2) 辨識作物異常偵測，利用攝相機進行影像拍攝蕃茄葉子其觀察病蟲害程度。使得果農能夠快速精準地找出熟成蕃茄與作物異常，大幅降低傳統人力花費大量時間巡視蕃茄園。針對蕃茄熟成度與作物異常，導入蕃茄熟成度與作物病蟲害偵測系統，將節省以往人工巡視浪費的人力，作物異常危機時會立即發出通知給相關人員，以利相關人員能立即處理，在降低人力成本的同時亦能加快發現時間，以至減少蕃茄受病蟲害危害受損風險，並且達到精準噴灑農藥。關鍵字：影像辨識、影像處理、蕃茄

關鍵字

影像辨識；影像處理；蕃茄

並列摘要

Student ID：M10756004 Title of thesis：：A Detection System of Tomato Maturity and Crop Anomaly Based on Image Recognition Technology Total page：51 Name of Institute：National Pingtung University of Science and Technology Graduate Institute of Management Information Systems Graduate date：January, 2021 Degree Conferred：Master Name of student：Wei-Zhi Zhang Adviser：Dr.Yuh-Jiuan Tsay The contents of abstract in this thesis： This research is based on image recognition technology to design and build a set of "Tomato maturity and crop pest detection system based on image recognition". The main purpose is: (1) To identify the maturity status of tomatoes and use a camera to capture tomatoes which is ripe or not. (2) Identify the unusual detection of crop, use the camera to take images of tomato leaves to observe the level of diseases and pests. This technology can help the farmer exactly find the irregular things and the maturity, which greatly reduces the time spent on traditional labo patrolling tomato gardens. Targeted at tomato maturity and crop abnormalities, the introduction of tomato maturity and crop pest detection systems will save the labor wasted by manual inspections in the past. When the crop is abnormal, the staff will be notified immediately so that they can deal with it immediately. The labor cost can also speed up the discovery time,to reduce the risk of damage on tomatoes by pests and diseases, and achieve precise pesticide spraying. Keywords：Image recognition, image processing, tomato