碩士 / 國立屏東科技大學 / 機械工程系所 / 105 / This study is mainly designed to apply the mechine vision to undertake the automatic counting of the shrimp larva and with such techniques as the automatic design, the mechanic vision and the image processing of the machine, we develop one set of the counting system with high efficiency, optimal precision and excellent stability.
The structure of this system can be classified as: the delivery mould, the detection mould as well as the self-developed counting software. The process of this system is: 1. placing the shrimp larva: put by hands the shrimp larva on the carrier of the delivery mould to conduct the delivery; 2. imaging in a dynamic state: move to the position of the advanced switch beneath the detection mould to go on with the imaging by sensor; 3. counting image: after the completion of the image prior processing, use the Blob to analyze and calculate the quantity; 4. collecting the larva: finally dump to the collecting bucket to complete the counting.
This study is undertaken to analyze 20, 30 and 40 Neocaridina denticulate sinensis var. “red” shrimps. The study results suggest that the highest precision rate as 97% for the 20 shrimp larva. As the quantity of the shrimp larva increases, the overlap rate raises up so as to cause the reduction of the precision rate to 77.25%. The counting quantity per hour for this system is approximately 36,000 shrimps and its precision rate is about 97%.
Identifer | oai:union.ndltd.org:TW/105NPUS5489012 |
Date | January 2017 |
Creators | Hsieh, Chun-Yu, 謝竣宇 |
Contributors | Lin, Yi-Hong, 林宜弘 |
Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
Language | zh-TW |
Detected Language | English |
Type | 學位論文 ; thesis |
Format | 53 |
Page generated in 0.0104 seconds