@article{oai:tuis.repo.nii.ac.jp:00000578, author = {伊丹, 伸 and 福田, 耕治 and 杉野, 隆三郎 and 三宅, 修平 and Itami, Shin and Fukuda, Koji and Sugino, Ryuzaburo and Miyake, Shuhei}, issue = {1}, journal = {東京情報大学研究論集}, month = {Sep}, note = {生物の複雑な行動を定量化する手法の確立は,数理情報学の産業応用上重要な課題である.本研究では,カオス・フラクタル解析を用いて,光刺激に対する魚群の遊泳行動の定量化を試みる.試験魚は水産学上有益性の高いマアジとし,LEDによる光刺激は赤色・緑色・青色・白色と4種類の波長を与え,異なる光刺激に対する魚群行動パターンの相異をみるために実験を行った.水槽実験から取得した動画像データより,個々のマアジの遊泳軌跡を抽出して時系列データに変換,その群行動パターンの特徴量として最大リアプノフ指数とHiguchi法によるフラクタル次元を算出した.その結果,魚群行動の定量化に成功し,特に赤色時の遊泳行動にカオス・フラクタルによる特徴量の大きさの違いが確認できたことを報告する., Establishment of the technique to quantify the complicated behaviors of creatures is an important subject for industrial applications of the mathematics and information science. In this study, we tried to quantify offish schooling behavior under light stimulation using chaos and fractal analysis. The experimental fish wasthe Japanese Jack Mackerel that has the high benefit in fishery science, and the light stimulation by LED was experimented giving light source such as red, green, blue and white one. We extracted a swimming trajectory from motion image data and calculated the maximum Lyapunov exponent and the fractal dimension by the Higuchi method. We used these chaos and fractal exponents as amount of characteristics to make a decision of the fish schooling behavior patterns. In our obtained results, we can show the effectiveness of chaos and fractal analysis for the quantification of fish schooling in which we can recognize the big amount of difference of red lighting among the other light source regarding the complicated swimming behavior in a water tank., 4, P}, pages = {41--50}, title = {カオス・フラクタル解析による光刺激に対する魚群行動の定量化への試み}, volume = {21}, year = {2017}, yomi = {イタミ, シン and フクダ, コウジ and スギノ, リュウザブロウ and ミヤケ, シュウヘイ} }