中文摘要：

嫁接可以有效地解决甜瓜连作障碍,提高产量,但嫁接作业劳动强度大、生产效率较低。目前,半自动化蔬菜嫁接机器人仍然需要人工对嫁接后的接缝进行质量检测,由于个人的主观性和技术熟练程度的不同,会影响嫁接质量和嫁接效率,从而影响嫁接成活率。因此,在嫁接之后,需要对嫁接苗进行接缝质量的自动识别与检测。为此,确定了嫁接幼苗接缝的成功标准,并利用图像采集技术和分割处理的基本理论与方法,研制嫁接幼苗接缝自动识别、检测的算法,完成了接缝检测软件程序编制。对嫁接幼苗接缝视觉检测识别系统软硬件进行了调试和试验研究,选取了60株甜瓜嫁接幼苗进行接缝图像检测和识别试验。试验结果表明:甜瓜嫁接幼苗接缝检测的成功率为90%,因茎秆折断弯曲造成了嫁接幼苗接缝的检测失败。

英文摘要：

Grafting can solve the problem of continuous cropping obstacle effectively and increase yield of the muskmel -on , but the labor intensity is large during grafting operation and the productivity is low .Grafting machines have been de-veloped with the widely application of grafting technology .The working process including detection the quality of gap after grafting is still made by manual , so the difference between the subjectivity of the individual and the level of technical pro -ficiency may affect the quality , efficiency of grafting and the survival rate .As a result , it is necessary to inspect grafting seedlings ’ gap automatically after grafting .The success criteria of recognizing grafting seedlings ’ gap were determined based on the theory and methods of image segmentation and the technology of image acquisition .The algorithm of auto-matic recognition and detection for grafting seedlings ’ gap was designed , and then the software programming was comple-ted and debugged .Experimental research on visual recognition and detection of grafting seedlings ’ gap for hardware and software system of grafting seedlings ’ gap was carried out with 60 muskmelon grafting seedlings .The results show that success rate reaches 90%, because the broken or bending of stem lead to the failure of test .The results meet the require-ments of the design .