适用于水质污染判别的鱼体尾频检测模型
|
摘要
为了实时准确地检测视频中运动鱼体的尾频特征,提出基于鱼体关键特征点的尾频检测模型。通过计算机视觉图像处理算法得到鱼体的二值图像,根据鱼体形体特征及摆尾特征,鱼体质心、尾点、头点等特征点坐标建立鱼体摆尾角度几何模型。最后通过对连续视频帧内鱼体摆尾角度分析,得到鱼体摆尾频率的变化规律。实验结果表明,该模型能够实时有效地获取单帧图像中鱼体摆尾特征,且在p H=7(水质正常)、p H=5(水质异常)、p H=3(水质恶化)三种水质中的鱼体尾频测试结果与实验观测到的尾频变化规律相符合,处理每帧图像平均耗时93.6 ms,满足系统实时性要求。
A fish tail frequency detection model based on critical feature points of fish body is proposed to detect the tail frequency characteristics of moving fish in video accurately in real-time. The binary image of the fish is obtained by means of computer vision image processing algorithm,and then the geometric model of the fish-tailing angle is established according to the physical and tail characteristics of fish,and feature points coordinates of fish centroid,tail point and head point. The fishtailing angles within the continuous video frames are analyzed to obtain the change rule of the fish-tailing frequency. The experimental results show that the model can effectively obtain the fish-tailing characteristics in a single image in real time,the test results of fish tail frequency are consistent with the observed change rules observed in experiment under the conditions of p H=7(normal water quality),p H=5(anomaly water quality)and p H=3(deteriorated water quality),and the average processing time of each frame image is 93.6 ms,which meets the real-time requirement of the system.
A fish tail frequency detection model based on critical feature points of fish body is proposed to detect the tail frequency characteristics of moving fish in video accurately in real-time. The binary image of the fish is obtained by means of computer vision image processing algorithm,and then the geometric model of the fish-tailing angle is established according to the physical and tail characteristics of fish,and feature points coordinates of fish centroid,tail point and head point. The fishtailing angles within the continuous video frames are analyzed to obtain the change rule of the fish-tailing frequency. The experimental results show that the model can effectively obtain the fish-tailing characteristics in a single image in real time,the test results of fish tail frequency are consistent with the observed change rules observed in experiment under the conditions of p H=7(normal water quality),p H=5(anomaly water quality)and p H=3(deteriorated water quality),and the average processing time of each frame image is 93.6 ms,which meets the real-time requirement of the system.
引文
[1]赵小强,彭红梅,高强.一种鱼体运动特征视频图像检测方法[J].西安邮电大学学报,2017,22(2):38-43.ZHAO Xiaoqiang,PENG Hongmei,GAO Qiang.A video image detection method for fish motion characteristics[J].Journal of Xi’an University of Posts and Telecommunications,2017,22(2):38-43.
[2]WANG Y H,CHEN J,TAO S F.Wavelet fusion application in water quality warning based on bio-detection technology[J].Applied mechanics&materials,2013,295:924-927.
[3]JEON W,KANG S H,LEEM J B,et al.Characterization of fish schooling behavior with different numbers of Medaka(Oryzias latipes)and goldfish(Carassius auratus)using a hidden Markov model[J].Physica A:statistical mechanics&its applications,2013,392(10):2426-2433.
[4]TAN J S,CHANG T K,OOI P L,et al.Fast and robust zebrafish segmentation and detection algorithm under different spectrum conditions[C]//2014 IEEE Sensor Applications Symposium.Queenstown,New Zealand:IEEE,2014:189-194.
[5]彭红梅.基于生物水质监测的鱼体运动状态检测系统研究[D].西安:西安邮电大学,2017.PENG Hongmei.Research on fish movement state detection system based on biological water quality monitoring[D].Xi’an:Journal of Xi’an University of Posts and Telecommunications,2017.
[6]程淑红,蔡菁,胡春海.基于视频算法的鱼类运动跟踪研究[J].光电工程,2011,38(2):14-18.CHENG S H,CAI J,HU C H.Fish motion tracking research based on video algorithm[J].Opto-electronic engineering,2011,38(2):14-18.
[7]VASSILISM P,IOANNISE P,FANI L,et al.A computer-vision system and methodology for the analysis of fish behavior[J].Aquacultural engineering,2012,46(1):53-59.
[8]陈久军,肖刚,应晓芳,等.鱼体尾频运动模型研究[J].中国图象图形学报,2009,14(10):2177-2180.CHEN J J,XIAO G,YING X F,et al.Fish activity model based on tail swing frequency[J].Journal of image&graphics,2009,14(10):2177-2180.
[9]程淑红,胡春海,蒋振洲,等.动态图像序列中鱼体体干运动建模[J].光电工程,2012,39(3):125-129.CHENG S H,HU C H,JIANG Z Z,et al.The motion model for fish′s soma in dynamic image sequences[J].Opto-electronic engineering,2012,39(3):125-129.
[10]张融,郑宏远,李录,等.一种计量青鳉鱼胸鳍和尾鳍摆动频率和幅值的计算机视觉算法[J].生态毒理学报,2015,10(4):154-161.ZHANG Rong,ZHENG Hongyuan,LI Lu,et al.A computer vision algorithm which was used for measuring the oscillation frequency of the Japanese Medaka′s Pectoral fin and Caudal fin[J].Asian journal of ecotoxicology,2015,10(4):154-161.
[11]周振宇,邵振洲,施智平,等.基于机器视觉的鱼类模式生物在线监测技术方法研究[J].生态毒理学报,2016,11(1):217-224.ZHOU Zhenyu,SHAO Zhenzhou,SHI Zhiping,et al.Study on the method of fish model organism on-line monitoring technology based on machine vision[J].Asian journal of ecotoxicology,2016,11(1):217-224.
[12]MING J,SHU Z,HE X Y.Instantaneous three-dimensional profile measurement of tail fin of carp[J].Optics&precision engineering,2009,17(9):2165-2169.
[13]胡江龙,方景龙,王大全.多目标跟踪算法在水质监测中的应用[J].机电工程,2012,29(5):613-615.HU J L,FANG J L,WANG D Q.Water quality monitoring using multi-object tracking algorithm[J].Journal of mechanical&electrical engineering,2012,29(5):613-615.
[14]WANG H,LIANG Y,WANG Z.Otsu image threshold segmentation method based on new genetic algorithm[J].Laser technology,2014,38(3):364-367.
[2]WANG Y H,CHEN J,TAO S F.Wavelet fusion application in water quality warning based on bio-detection technology[J].Applied mechanics&materials,2013,295:924-927.
[3]JEON W,KANG S H,LEEM J B,et al.Characterization of fish schooling behavior with different numbers of Medaka(Oryzias latipes)and goldfish(Carassius auratus)using a hidden Markov model[J].Physica A:statistical mechanics&its applications,2013,392(10):2426-2433.
[4]TAN J S,CHANG T K,OOI P L,et al.Fast and robust zebrafish segmentation and detection algorithm under different spectrum conditions[C]//2014 IEEE Sensor Applications Symposium.Queenstown,New Zealand:IEEE,2014:189-194.
[5]彭红梅.基于生物水质监测的鱼体运动状态检测系统研究[D].西安:西安邮电大学,2017.PENG Hongmei.Research on fish movement state detection system based on biological water quality monitoring[D].Xi’an:Journal of Xi’an University of Posts and Telecommunications,2017.
[6]程淑红,蔡菁,胡春海.基于视频算法的鱼类运动跟踪研究[J].光电工程,2011,38(2):14-18.CHENG S H,CAI J,HU C H.Fish motion tracking research based on video algorithm[J].Opto-electronic engineering,2011,38(2):14-18.
[7]VASSILISM P,IOANNISE P,FANI L,et al.A computer-vision system and methodology for the analysis of fish behavior[J].Aquacultural engineering,2012,46(1):53-59.
[8]陈久军,肖刚,应晓芳,等.鱼体尾频运动模型研究[J].中国图象图形学报,2009,14(10):2177-2180.CHEN J J,XIAO G,YING X F,et al.Fish activity model based on tail swing frequency[J].Journal of image&graphics,2009,14(10):2177-2180.
[9]程淑红,胡春海,蒋振洲,等.动态图像序列中鱼体体干运动建模[J].光电工程,2012,39(3):125-129.CHENG S H,HU C H,JIANG Z Z,et al.The motion model for fish′s soma in dynamic image sequences[J].Opto-electronic engineering,2012,39(3):125-129.
[10]张融,郑宏远,李录,等.一种计量青鳉鱼胸鳍和尾鳍摆动频率和幅值的计算机视觉算法[J].生态毒理学报,2015,10(4):154-161.ZHANG Rong,ZHENG Hongyuan,LI Lu,et al.A computer vision algorithm which was used for measuring the oscillation frequency of the Japanese Medaka′s Pectoral fin and Caudal fin[J].Asian journal of ecotoxicology,2015,10(4):154-161.
[11]周振宇,邵振洲,施智平,等.基于机器视觉的鱼类模式生物在线监测技术方法研究[J].生态毒理学报,2016,11(1):217-224.ZHOU Zhenyu,SHAO Zhenzhou,SHI Zhiping,et al.Study on the method of fish model organism on-line monitoring technology based on machine vision[J].Asian journal of ecotoxicology,2016,11(1):217-224.
[12]MING J,SHU Z,HE X Y.Instantaneous three-dimensional profile measurement of tail fin of carp[J].Optics&precision engineering,2009,17(9):2165-2169.
[13]胡江龙,方景龙,王大全.多目标跟踪算法在水质监测中的应用[J].机电工程,2012,29(5):613-615.HU J L,FANG J L,WANG D Q.Water quality monitoring using multi-object tracking algorithm[J].Journal of mechanical&electrical engineering,2012,29(5):613-615.
[14]WANG H,LIANG Y,WANG Z.Otsu image threshold segmentation method based on new genetic algorithm[J].Laser technology,2014,38(3):364-367.