自然场景下苹果采摘机器人视觉系统研究-《计算机技术与发展》

文章信息/Info

Author(s):: LI Na; CHEN Ning; Xi’an Polytechnic University，Xi’an 710048，China

Keywords:: ntelligent agriculture; robot; visual system; fruit identification; localization of target

摘要:: 随着现代农业智能化技术的飞速开展,智慧农业系统的应用越来越多.农业采摘机器人作为智慧农业的一个新兴领域,成为近年来研究的热点.农业机器人的一个重要组成部分—视觉系统,它的性能好坏将直接影响果实识别和目标定位的准确性,从而影响其采摘效率.面对自然环境中的果实目标,采摘机器人的视觉系统仍然面临着巨大的挑战.首先对智慧农业的研究现状进行分析概括,然后对目前自然场景中的苹果果实采摘现状进行分析,进一步对自然场景中的苹果果实受光照影响、果实重叠、果实相互遮挡、树枝遮挡等现状的识别以及所采用的研究方法进行了分析总结,指出各方面需要改进的问题,最后就自然场景中苹果果实目标的识别及其定位技术进行了展望.

Abstract:: With the rapid development of modem agricultural intelligent technology，the application of intelligent agricultural systemis more and more．As a new field of intelligent agriculture，agricultural picking robot has become a hot topic in recent years．As an imporrant part of agricultural robot，the performance of vision system will directly affect the accuracy of fruit recognition and target positioning，thus afecfing the efficiency of picking．Facing the fruit target in the natural environment，the visual system of picking robot still faces great challenges．We firstly summarize the wisdom agriculture research status．Then we analyze the present situation of apple fruit picking in the natural scene，further summarizing the identification of apple fruit under the influence of light，fruit overlapping，fruitmumal OC-clusion and tree occlusion in the natural scene and the research methods adopted，pointing out the problem tO be improved．Finally，we prospect the identification of fruittarget in natural scene and itspositioning technology．