摘要
在過去幾十年來，科學家一直想了解人類大腦神經網路的連結，但由於大腦的研究受到各國很嚴厲的限制，甚至禁止，所以科學家們選取果蠅當作研究對象，由於果蠅在一些行為表現與遺傳性的疾病跟人有相當的相似性，並且果蠅染色體數量只有八條，相對人類四十六條來說是一個較容易研究的目標，並且可藉由人工的方式調控基因，改變果蠅的行為能力，經由此調控機制認識果蠅腦內神經網路的連結方式，揭開腦科學這神秘的面紗。為了要觀察與分析果蠅的行為表現，於是有許多的研究團隊針對特定觀察的行為各自發展了專屬的觀測分析系統，也利用這些系統發現許多人類所不知道的腦科學知識。
本論文提出一套即時數位溫控與果蠅影像辨識系統，本系統能夠設定鋁板左右兩端的目標溫度，並利用Sensor得知鋁板目前的溫度，再由分析系統模組採用PID控制，對鋁板加熱或致冷，使鋁板維持在一個10℃~40℃的溫度梯度表現，也提供一個行為驗證平台，給予科學家作為果蠅腦神經網路的研究，經由人工基因改質之果蠅，會在溫度梯度10℃~40℃的鋁板上自由活動，實驗過程都在暗室中進行，果蠅經過三十分鐘的活動後，系統將會自動開燈拍攝一張果蠅在鋁板上的分佈影像，並利用影像辨識的方法記錄不同溫度範圍內果蠅的數量，經由程式繪出果蠅在鋁板上的分佈趨勢圖，同時搭配公式演算證明果蠅是否有冷溫或熱溫感知的缺陷，此系統擁有良好的溫度控制及導熱性，體積上也縮小許多，加入影像辨識功能，藉由影像辨識判斷出圖片內果蠅的位置分佈，此功能的加入勢必能減少人為誤判及免除表示鋁板溫度範圍的特定標線，對於腦科學神經網路研究來說，提供了一套精準且方便的系統。

Abstract
Over the past few decades, scientists were seeking to understand the neural network of a human brain. Due to severe restricted set by the government across the globe, including prohibition on experimenting human subjects. Drosophila Melanogaster has been chosen as a tool by the scientists, due to its similarities in social behavior and hereditary diseases posed by a human. Having only 4-pairs of chromosomes makes it much easier to study as, compared with 23-pairs possessed by a homo sapien. There exists ample of space to manipulate gene regulation by artificial means, for understanding the control mechanisms by linking that to a neural network of Drosophila. In order to understand the fly behavior, several teams have developed customized set of tools to study drosophila’s subtle behaviors. Those tools were able to unveil several mysteries for the knowledge base of brain science.
Here is this thesis we present a ”Development of Drosophila Melanogaster System for Image Identification and Real-time Temperature Control”. The system consist of thermal grating aluminum plane, controlled by a PID control module and sensors attached to the system in capable of displaying the temperature gradient. This system is quite flexible in terms of controlling the temperature various from 10℃~40℃ across the aluminum plane, which will enable scientists working on drosophila neural network a handy tool to work on. We design a set experiment in which the lowest temperature is set at 10℃ at one end of the aluminum plane and gradually increasing the temp up to 40℃ towards other end of the aluminum plane and clusters of drosophila flies were let to move freely on the temperature gradient aluminum surface for 30 min in dark. At the end of 30min the system automatically turn on light and capture the image of the surface which is able to determine the distribution of flies on the aluminum plane with respect to temperature gradient with the help of image recognition technique. Using this system we were able to use a proven formula of cold score and warm score calculation for drosophila, for investigating the defects suffered by drosophila flies. This system has a very good temperature control and thermal stability along with image recognition capability and automatic analysis of fly distribution.

參考文獻
[1] Wolf, F.W. and Heberlein, U. Invertebrate models of drug abuse. J. Neurobiol. 54,
161–178 (2003).
[2] Guarnieri, D.J. & Heberlein, U. Drosophila melanogaster, a genetic model system
for alcohol research. Int. Rev. Neurobiol. 54, 199–228 (2003).
[3] Ho, K.S. & Sehgal, A. Drosophila melanogaster: an insect model for fundamental
studies of sleep. Methods Enzymol. 393, 772–793 (2005).
[4] Shaw, P., Ocorr, K., Bodmer, R. & Oldham, S. Drosophila aging. Exp. Gerontol. 43, 5–10 (2008).
[5] Konsolaki, M., Song, H.J., Dobbs, W. & Garza, D. P2–109 Drosophila models of Alzheimer’s-related pathways. Neurobiol. Aging 25, S255–S255 (2004).
[6] Sung-Tae Hong, Sunhoe Bang, Donggi Paik, Jongkyun Kang, Seungyoon Hwang, Keunhye Jeon, Bumkoo Chun, Seogang Hyun, Youngseok Lee, and Jaeseob Kim, Histamine and its receptors modulate temperature preference behaviors in Drosophila. The Journal of Neuroscience, 26(27):7245–7256(2006).
[7] Roseanna B. Ramazani, Harish R. Krishnan, Susan E. Bergeson, Nigel S. Atkinson, Computer automated movement detection for the analysis of behavior. Journal of Neuroscience Methods 162 171–179(2007).
[8] John E. Zimmerman, David M. Raizen, Matthew H. Maycock, Greg Maishlin, Allan I. Pack. A video method to study Drosophila sleep. SLEEP. Vol. 31, No. 11(2008).
[9] Kristin Branson, Alice A Robie, John Bender, Pietro Perona & Michael H Dickinson. High-throughput ethomics in large groups of Drosophila. Nature method(2009).
[10] Jasper C. Simon, Michael H. Dickinson, A New chamber for studying the behavior of Drosophila. Plos One Volume 5 Issue 1(2010).
[11] 劉宗和，“研究果蠅行為的即時影像分析與雷射刺激之系統研發”，國立清華大學動力機械工程學系研究所碩士論文，2010年
[12] 吳明親，“即時果蠅行為追跡及分析訓練系統”，國立清華大學動力機械工程學系研究所博士論文計畫書（初稿），2010年