城市水体表观污染颜色识别技术的优化研究

城市水体表观污染指数是评价和表征城市水体感官污染程度的重要指标。在基于吸收光谱和颜色识别技术的原有检测方法基础上,以苏州城区景观水体为例,通过实地采样与检测,对人工判别与颜色识别程序的判别结果进行了对比验证。根据验证结果分别从颜色判别模型、颜色识别程序参数的调整两个方面对颜色识别技术进行优化,最终验证吻合率均上升到90%以上,让识别更精准。同时验证使用手机照片进行颜色判别的吻合率,灰色水体吻合率达到100%,黄色水体吻合率达到96.7%,绿色水体吻合率达到100%,使识别更加便捷,为城市水体表观污染现场检测设备和线上检测系统的研发、使用与推广提供技术支持。

The apparent pollution index of urban water body is an important indicator for evaluating and characterizing the degree of sensory pollution of urban water body.Based on the original detection methods of color recognition technology and absorption spectrum,the research team took the river course in Suzhou urban area as an example,and through field sampling and detection,compared and verified the identification results of manual identification and color recognition procedures.According to the verification results,the color recognition technology was optimized from two aspects,namely,the color discrimination model and the adjustment of color recognition program parameters.The final verification coincidence rate increased to more than 90%,making the recognition more accurate.At the same time,the matching rate of color discrimination using mobile phone photos is verified,the matching rate of gray water bodies is 100%,the matching rate of yellow water body is 96.7%,and the matching rate of green water body is 100%,making the recognition more convenient,and providing technical support for the development,use and promotion of on-site detection equipment and online detection system for urban water surface pollution.