《DSP论文：基于DSP的水稻杂草识别研究》由会员分享，可在线阅读，更多相关《DSP论文：基于DSP的水稻杂草识别研究（3页珍藏版）》请在金锄头文库上搜索。
1、DSP论文：基于DSP的水稻杂草识别研究【中文摘要】水稻杂草对水稻生长带来极大危害，它是长期适应 水稻耕种、气候、土壤等生态环境而生存下来的，具有很强的适应能 力。杂草与水稻争夺养分、水分及光照等，易于助长病虫害的滋生和 蔓延，降低水稻产量和质量，给水稻粮食生产造成巨大损失。面对严重 的水稻杂草危害，目前主要的除草方式有人工除草和化学除草。人工 除草需要大量劳动力，效率低；化学除草由于高效的除草能力而备受 广大农民的青睐，成为目前最主要的水稻除草方式。化学除草一般采 用大面积喷洒农药，这种喷洒方式不仅提高了农业成本，而且破坏了 土壤质量,污染了环境，不利于农业的可持续发展。大量实验表明水稻 田
2、间杂草分布是不均匀的，因此要研究一种变量喷洒的方法，即在有 杂草的地方喷药，在没有杂草或杂草密度低的地方停止喷药。为实现 变量喷药,首先要实现对水稻田间杂草的实时识别。本研究以DSP为主处理芯片，搭建水稻杂草识别系统，实现对水稻杂草的实时识别与 标记,为后续定点变量喷药除草系统的开发奠定基础。本论文主要研 究内容、结果与创新点包括：（1）选用TI公司的TMS320DM64为主处 理芯片,搭建水稻杂草识别硬件平台，完成水稻杂草识别算法编程，实 现系统对水稻杂草的实时识别与标记。（2）研究提取水稻田间图像感 兴趣区域方法，通过对各颜色特征的分析，得出2g-r-b、2G-R-B和色 度H三种颜色特征
3、值都可以用来进行水稻田间背景的分割。对于水 稻、杂草与背景的分割，提出结合超绿和超红颜色因子组合,应用最大类间方差法实现对水稻田间背景的自动分割；(3)研究水稻田间图像 采集方法,提出针对此采集方式的杂草区域分割算法，实现对水稻田 间杂草区域的识别，并完成对杂草区域的区域标记和面积计算；(4) 在对五种水稻田间杂草(稗草(Echinochloaphyllopogon Koss)、矮慈 菇(Sagittaria Pygmaea Miq)、丁香蓼(Ludwigia prostrata Roxb) 、四叶萍(Marsilea quadrifolia)、鸭舌草(Monochoria vaginalis
4、Presl)的特征提取上，利用非均匀量化的直方图提取颜色特征，提高 系统识别鲁棒性。利用灰度共生矩阵提取纹理特征，压缩了杂草图像的灰度级,减少了 3/4的计算量，降低杂草特征提取的复杂度；在对五 种常见水稻田间杂草的种类识别种，分别应用贝叶斯、支持向量机以 及BP神经网络三种分类器对颜色参数集、形态参数集、纹理参数集 以及所有参数集进行了种类识别研究，结果表明,BP神经网络在水稻 田间杂草种类的识别中更具优越性；【英文摘要】 The rice weeds bri ng great harm to the growthof rice; they have strong survivability
5、 as they had adapt the rice farming, climate, soil and environment. They compete for nutrients,water, light with rice, and weeds in rice field areeasy to con tribute to the spread of pests and diseases, the yield and quality of rice are great reduced, caused huge losses to rice.Face the serious dama
6、ge of weeds in rice field, the main ly methods are manual weed ing and chemical weed ing atpresent. The manual weeding is low efficiency,labor-intensive, because the chemical weed can effectively con trol the weeds,it has becomethe mainly wayof weeding in rice field. At present the method to weed is
7、 to spray the herbicide, and the means to spray is the well-distributed spray ing. This method, not only improve the cost of agriculture but also damagethe quality of field and pollute air. Study indicates that weed isn twell-distributed, so we should find a variable-controlling spay ing method. It
8、is to spray in the regi on where there are weed patches, and to stop in the region where there are not.In order to do so, the first thi ng is to realize the detecti on of weed patches. In this paper we selected the DSP as the main process ing chip, con struct the rice weed recog niti on system, and
9、realized the reorganizationof rice weedregion, marked theweed region in real time, and laid a foundation for the subsequent development of automatic variable spraying system. The main content and results are as follows:(1)Select TI sTM320DM642 DSP as the main process ing chip, con structed the hardware system of rice weed identification,and programmedtherice weed identification algorithm,implemented the rice weedidentification and marking in real time.(2)Rearched the methods of extracting the region of interest on the images of rice field, and the color image an alysis (2g-r-b), (2G-R-B),
