摘要
为了提高湖面垃圾识别的准确性,降低对非垃圾区域的误判,提出了一种基于HSV颜色空间模型和垃圾特征检测的湖面垃圾视觉识别算法。该算法首先在HSV颜色空间中找出图像中疑似垃圾的区域;再利用垃圾的特征进行初步的垃圾识别;最后通过背景过滤和形态学特征分析,进行最终确认。实验表明,该算法具有较高的湖面垃圾检测率。
In order to improve the accuracy of lake garbage identification and reduce the misjudgment of non-garbage areas,a lake garbage visual recognition algorithm based on HSV color space model and garbage feature detection is proposed in this paper.The algorithm first finds out the area of suspected garbage in the image in the HSV color space,then uses the characteristics of the garbage for preliminary garbage identification.Finally,it conducts final confirmation through background filtering and morphological feature analysis.Experiments show that the algorithm has a high detection rate of lake garbage.
出处
《工业控制计算机》
2021年第1期77-78,127,共3页
Industrial Control Computer
基金
2019年国家级大学生创新创业训练计划项目(201911819004)
2020年广东省科技创新战略专项资金资助立项项目(pdjh2020b0578)。
关键词
机器视觉
湖面垃圾识别
过滤背景
特征识别
machine vision
lake garbage recognition
filter background
feature recognition