渔业资源声学评估以其无选择性、覆盖面积大、快速准确、不伤资源等优点被应用到多种水生生物的资源评估工作中。但是,在使用回声积分系统采集水体声学信号时,由于受声学仪器、调查船只、周围环境、混在生物等因素影响,会产生很多噪声和混响等干扰信号。在使用声学数据进行资源评估之前,应将数据中的噪声和混响进行剔除,避免对资源量评估产生影响。文章将国内外所使用的各种噪声剔除方法归纳总结为三类:设定积分阈值法、科学渔探仪被动模式噪声剔除法和科学渔探仪主动模式噪声剔除法,并指出了这些方法的优缺点,希望对渔业资源声学评估的数据后处理过程有所帮助。
With the features of non-selection, wide coverage, high accuracy and non-damage, fishery acoustics is widely used in resource assessment of various kinds of aquatic organisms. But due to instrumental, environmental, biological, and artificial reasons, fishery acoustics is easily affected by interference signals, such as noise and reverberation. In some cases, the effects of noises are so obvious that we have to remove them before calculating the resource. The paper summarized various methods used to remove noises at home and abroad to three categories: integration thresholds, background noise removal technique based on passive mode sonar data and background noise removal technique based on active mode sonar data. And the paper pointed out the advantages and disadvantages of these methods, hoping that it could be helpful in post-processing of acoustic data.