發(fā)布時(shí)間：2019-05-17 19:00

【摘要】：菌落檢測(cè)是農(nóng)業(yè)、食品、醫(yī)藥衛(wèi)生等行業(yè)進(jìn)行質(zhì)量檢測(cè)的重要方法之一。依靠肉眼觀察的人工檢測(cè)方法速度慢、結(jié)果重復(fù)性差,而且工作繁重乏味,工作效率低。目前,市場(chǎng)上雖然出現(xiàn)了菌落自動(dòng)計(jì)數(shù)儀,但大都以PC機(jī)作為運(yùn)算平臺(tái),便攜性差,難以滿足實(shí)時(shí)檢測(cè)要求。近年來(lái)嵌入式技術(shù)飛速發(fā)展,嵌入式設(shè)備在運(yùn)算速度和存儲(chǔ)容量上都有了長(zhǎng)足的進(jìn)步,嵌入式操作系統(tǒng)的發(fā)展也日趨成熟。由于嵌入式設(shè)備本身具有便攜性好、成本低的特點(diǎn),使嵌入式技術(shù)得到了更為廣泛的應(yīng)用。 針對(duì)菌落檢測(cè)對(duì)實(shí)時(shí)性和便攜性的要求,本研究結(jié)合嵌入式技術(shù),設(shè)計(jì)了基于ARM平臺(tái)的嵌入式菌落智能計(jì)數(shù)系統(tǒng),并完成了原型樣機(jī)的開(kāi)發(fā)。系統(tǒng)以S3C6410作為核心處理器,采用廣角紅外CCD攝像頭作為視頻采集設(shè)備。為了提高嵌入式菌落智能計(jì)數(shù)系統(tǒng)的整體運(yùn)行效率,我們對(duì)系統(tǒng)使用的Windows CE 6.0內(nèi)核進(jìn)行了裁剪和定制;為了達(dá)到最好的檢測(cè)效果,本設(shè)計(jì)對(duì)樣機(jī)結(jié)構(gòu)和照明系統(tǒng)進(jìn)行了精心設(shè)計(jì)。 菌落自動(dòng)計(jì)數(shù)算法是整個(gè)系統(tǒng)性能優(yōu)劣的關(guān)鍵。我們使用OPENCV計(jì)算機(jī)視覺(jué)庫(kù)開(kāi)發(fā)了菌落智能計(jì)數(shù)算法,研究了一種基于霍夫變換的培養(yǎng)皿邊緣提取算法,取得了良好的效果。粘連菌落的分割是菌落自動(dòng)計(jì)數(shù)算法的難點(diǎn),本文提出了一種基于迭代腐蝕算法的菌落計(jì)數(shù)算法,并通過(guò)多次條件膨脹的方法消除了傳統(tǒng)迭代腐蝕算法對(duì)較大菌落的“過(guò)分割”問(wèn)題,使平均計(jì)數(shù)偏差控制在6%以內(nèi),取得了很好的效果。最后,本文完成了算法的移植和PC演示版軟件的開(kāi)發(fā),并最終實(shí)現(xiàn)了便攜式菌落智能計(jì)數(shù)系統(tǒng)的開(kāi)發(fā)和測(cè)試工作; 總之,本研究對(duì)菌落自動(dòng)計(jì)數(shù)算法進(jìn)行了有意義的探索和嘗試,并在此基礎(chǔ)上開(kāi)發(fā)了嵌入式樣機(jī)系統(tǒng),研究成果具有一定的理論意義和應(yīng)用價(jià)值。
[Abstract]:Colony detection is one of the important methods for quality testing in agriculture, food, medicine and hygiene. The manual detection method based on naked eye observation has the advantages of slow speed, poor repeatability, tedious work and low work efficiency. At present, although colony automatic counter appears in the market, most of them use PC as the operation platform, and the portability is poor, so it is difficult to meet the requirements of real-time detection. With the rapid development of embedded technology in recent years, embedded devices have made great progress in computing speed and storage capacity, and the development of embedded operating system is becoming more and more mature. Because the embedded device itself has the characteristics of good portability and low cost, embedded technology has been more widely used. In order to meet the requirements of real-time and portability of colony detection, an embedded colony intelligent counting system based on ARM platform is designed based on embedded technology, and the prototype is developed. The system uses S3C6410 as the core processor and wide-angle infrared CCD camera as the video capture equipment. In order to improve the overall operation efficiency of the embedded colony intelligent counting system, we cut and customize the Windows CE 6.0 kernel used in the system. In order to achieve the best detection effect, the prototype structure and lighting system are carefully designed. Colony automatic counting algorithm is the key to the performance of the whole system. An intelligent colony counting algorithm is developed by using OPENCV computer visual library, and a petri dish edge extraction algorithm based on Hoff transform is studied, and good results are obtained. The segmentation of adhesive colonies is a difficult point in colony automatic counting algorithm. In this paper, a colony counting algorithm based on iterative corrosion algorithm is proposed. Through the method of multiple conditional expansion, the problem of "over-segmentation" of large colonies by traditional iterative corrosion algorithm is eliminated, and the average counting deviation is controlled within 6%, and good results are obtained. Finally, this paper completes the transplantation of the algorithm and the development of PC demonstration software, and finally realizes the development and testing of portable colony intelligent counting system. In a word, this study has carried on the meaningful exploration and the attempt to the colony automatic counting algorithm, and has developed the embedded prototype system on this basis, the research result has certain theoretical significance and the application value.
【學(xué)位授予單位】：天津大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2012
【分類號(hào)】：R318.6

【參考文獻(xiàn)】

相關(guān)期刊論文 前6條

1 宋強(qiáng),徐科,徐金梧,孫浩,王金華,王春梅;基于圖象處理的棒材自動(dòng)計(jì)數(shù)技術(shù)[J];鋼鐵;2004年05期

2 周瑩莉,曾立波,劉均堂,余曉敏;基于圖像處理的菌落自動(dòng)計(jì)數(shù)方法及其實(shí)現(xiàn)[J];數(shù)據(jù)采集與處理;2003年04期

3 孫明,王一鳴,凌云,張小超;基于色調(diào)的黃粒米檢測(cè)方法[J];農(nóng)業(yè)機(jī)械學(xué)報(bào);2005年08期

4 凌云,王一鳴,孫明,孫紅,張小超;基于機(jī)器視覺(jué)的大米外觀品質(zhì)檢測(cè)裝置[J];農(nóng)業(yè)機(jī)械學(xué)報(bào);2005年09期

5 侯彩云,王一鳴,凌云,孫劍鋒,孫明,賈貴儒,林夕;堊白米粒的計(jì)算機(jī)圖像識(shí)別[J];農(nóng)業(yè)工程學(xué)報(bào);2002年03期

6 尤育賽,于慧敏;一種重疊紅細(xì)胞圖像的分離方法[J];中國(guó)圖象圖形學(xué)報(bào);2005年06期

相關(guān)博士學(xué)位論文 前1條

1 劉相濱;類圓性顆粒圖像分割技術(shù)研究[D];湖南大學(xué);2006年

，

本文編號(hào)：2479334