学科分类
/ 1
1 个结果
  • 简介:Fishfindershavealreadybeenwidelyavailableinthefishingmarketforanumberofyears.However,thesizesofthesefishfindersaretoobigandtheirpricesareexpensivetosuitfortheresearchofroboticfishormini-submarine.Thegoalofthisresearchistoproposealow-costfishdetectorandclassifierwhichsuitsforunderwaterrobotorsubmarineasaproximitysensor.Withsomepre-conditioninhardwareandalgorithms,theexperimentalresultsshowthattheproposeddesignhasgoodper-formance,withadetectionrateof100%andaclassificationrateof94%.Boththeexistingtypeoffishandthegroupbehaviorcanberevealedbystatisticalinterpretationssuchashoveringpassionandsparseswimmingmode.

  • 标签: FISH detection classincation artificial NEURAL network