自動(dòng)水產(chǎn)養(yǎng)殖作業(yè)船的視覺(jué)導(dǎo)航技術(shù)研究.doc
約38頁(yè)DOC格式手機(jī)打開(kāi)展開(kāi)
自動(dòng)水產(chǎn)養(yǎng)殖作業(yè)船的視覺(jué)導(dǎo)航技術(shù)研究, 1.85萬(wàn)字自己原創(chuàng)的畢業(yè)論文，僅在本站獨(dú)家出售，重復(fù)率低，推薦下載使用 摘要 數(shù)字圖像處理就是利用計(jì)算機(jī)對(duì)圖像信息進(jìn)行加工，以滿足人的視覺(jué)心理或者應(yīng)用的需求。圖像識(shí)別所討論的問(wèn)題，是研究用計(jì)算機(jī)代替人自動(dòng)地處理大量的物理信息，解決人類所不能識(shí)別的問(wèn)題。對(duì)于計(jì)算機(jī)來(lái)說(shuō)，在實(shí)際工作環(huán)...
內(nèi)容介紹
此文檔由會(huì)員 淘寶大夢(mèng) 發(fā)布
自動(dòng)水產(chǎn)養(yǎng)殖作業(yè)船的視覺(jué)導(dǎo)航技術(shù)研究
1.85萬(wàn)字
自己原創(chuàng)的畢業(yè)論文，僅在本站獨(dú)家出售，重復(fù)率低，推薦下載使用
摘要 數(shù)字圖像處理就是利用計(jì)算機(jī)對(duì)圖像信息進(jìn)行加工，以滿足人的視覺(jué)心理或者應(yīng)用的需求。圖像識(shí)別所討論的問(wèn)題，是研究用計(jì)算機(jī)代替人自動(dòng)地處理大量的物理信息，解決人類所不能識(shí)別的問(wèn)題。對(duì)于計(jì)算機(jī)來(lái)說(shuō)，在實(shí)際工作環(huán)境里，圖像場(chǎng)景已有較大的變化。因此要區(qū)分圖像屬于哪一類，往往要通過(guò)一系列關(guān)鍵技術(shù)來(lái)實(shí)現(xiàn)。由此產(chǎn)生的圖像識(shí)別方法也有很多。
本文以水面障礙物目標(biāo)（竹竿、漁網(wǎng)等）的圖像識(shí)別為例，從白天自然光條件下的圖像識(shí)別方法入手，展開(kāi)研究。本文以數(shù)學(xué)圖像處理為重點(diǎn)，較詳細(xì)地論述了對(duì)水面上障礙物目標(biāo)進(jìn)行圖像識(shí)別的技術(shù)和過(guò)程。由于采集到的圖像目標(biāo)和背景灰度色差不是很明顯，對(duì)于的圖像分割不適合采用色差分割法，于是，本文采用了邊緣檢測(cè)的方法進(jìn)行識(shí)別。包括圖像預(yù)處理、邊緣提取、噪聲去除、圖像銳化和領(lǐng)域平均法處理等的主要過(guò)程。最終達(dá)到目標(biāo)識(shí)別的目的。研究結(jié)論如下：
（1）灰度化處理使原來(lái)的彩色圖像灰度化，并進(jìn)行圖像縮放以減少計(jì)算量，提高運(yùn)算效率，對(duì)灰度圖像進(jìn)行邊緣檢測(cè)，濾波和銳化，突出目標(biāo)，削弱背景噪聲。
（2）采用鄰域平均法對(duì)圖像中的目標(biāo)和遠(yuǎn)處水面背景噪聲進(jìn)行分割，提取出圖像中的目標(biāo)，為進(jìn)一步得到目標(biāo)的位置等信息打下基礎(chǔ)。
本文在較廣泛地調(diào)研文獻(xiàn)的基礎(chǔ)上，對(duì)圖像識(shí)別系統(tǒng)進(jìn)行了較為全面的綜述，并以較為大量文字和具體的實(shí)例，通過(guò)使用常用的仿真語(yǔ)言和軟件對(duì)基于數(shù)字圖像處理的障礙物的識(shí)別進(jìn)行了研究，獲得了較理想的識(shí)別結(jié)果。
關(guān)鍵詞： 圖像處理 水面目標(biāo)識(shí)別 邊緣檢測(cè) 視覺(jué)導(dǎo)航
Abstract Digital image processing is the use of computers for image information processing to meet the person's visual psychology or application requirements. Issues discussed at image recognition, is the study of automatically replace people use your computer to work with large amounts of physical information, solve the problem of human beings does not recognize. For the computer, in the actual working environment, image scene has a larger change. Therefore, to distinguish between images fall into which category, often through a series of key technology to achieve. The resulting image recognition also has many kinds of methods.
Based on the target surface obstacles (bamboo, fishing nets, etc.) of image recognition, for example, from the natural light during the day under the condition of image recognition method of study. Focusing on mathematical image processing, this paper discusses on the surface of the water obstacles target image recognition technology and process. Because of the collected image object and background gray color difference is not very obvious, for the image segmentation is not suitable for color segmentation method is used to, so, in this paper, the edge detection method was adopted for identification. Including image preprocessing, edge detection and noise removal and image sharpening and average method, the main process. Ultimately achieve the goal of target recognition. Research conclusions are as follows:
(1)Gray processing make the original color image gray level and image zooming to reduce the amount of calculation, improve the efficiency of operation, and the gray image edge detection, filtering and sharpening, highlight the objectives and weaken the background noise.
(2)The neighborhood average method is adopted to image the target and background noise in the distance the water division, extract the target in the image, the information such as the location of the target is obtained for the further lay the foundation.
In the research literature on the basis of extensively in this paper, the image recognition system was carried out, and with relatively large amounts of text and concrete examples, through the use of the commonly used language and software simulation of obstacle recognition based on digital image processing was studied, and obtained the ideal recognition results.
Key words Image processing Target identification Edge detection Visual navigation
目 錄
第一章 緒 論 1
1.1 研究的背景 1
1.2 本文的研究目的和意義 1
第二章 論文的研究背景 3
2.1 機(jī)器人視覺(jué)導(dǎo)航技術(shù)的應(yīng)用及展望 3
2.2 在工業(yè)檢測(cè)中的應(yīng)用 4
第三章 圖像識(shí)別系統(tǒng) 8
3.1 圖像預(yù)處理 8
3.1.1 灰度化和二值化 9
3.1.2 邊緣檢測(cè) 10
3.2 形態(tài)學(xué)處理 12
3.3 圖像的識(shí)別 14
3.4 小結(jié) 15
第四章 基于數(shù)字圖像處理的障礙物識(shí)別 16
4.1 仿真環(huán)境簡(jiǎn)介 16
4.2 MATLAB處理 18
4.2.1 灰度化和二值化 18
4.2.2 圖像數(shù)據(jù)處理 20
4.2.3 形態(tài)學(xué)處理 21
4.2.4 濾波處理 21
4.2.5 圖像銳化和目標(biāo)識(shí)別 23
4.3 小結(jié) 24
第五章 總結(jié)與展望 25
致 謝 26
參考文獻(xiàn) 27
附錄一（程序代碼） 29
附錄二（程序運(yùn)行結(jié)果） 31
1.85萬(wàn)字
自己原創(chuàng)的畢業(yè)論文，僅在本站獨(dú)家出售，重復(fù)率低，推薦下載使用
摘要 數(shù)字圖像處理就是利用計(jì)算機(jī)對(duì)圖像信息進(jìn)行加工，以滿足人的視覺(jué)心理或者應(yīng)用的需求。圖像識(shí)別所討論的問(wèn)題，是研究用計(jì)算機(jī)代替人自動(dòng)地處理大量的物理信息，解決人類所不能識(shí)別的問(wèn)題。對(duì)于計(jì)算機(jī)來(lái)說(shuō)，在實(shí)際工作環(huán)境里，圖像場(chǎng)景已有較大的變化。因此要區(qū)分圖像屬于哪一類，往往要通過(guò)一系列關(guān)鍵技術(shù)來(lái)實(shí)現(xiàn)。由此產(chǎn)生的圖像識(shí)別方法也有很多。
本文以水面障礙物目標(biāo)（竹竿、漁網(wǎng)等）的圖像識(shí)別為例，從白天自然光條件下的圖像識(shí)別方法入手，展開(kāi)研究。本文以數(shù)學(xué)圖像處理為重點(diǎn)，較詳細(xì)地論述了對(duì)水面上障礙物目標(biāo)進(jìn)行圖像識(shí)別的技術(shù)和過(guò)程。由于采集到的圖像目標(biāo)和背景灰度色差不是很明顯，對(duì)于的圖像分割不適合采用色差分割法，于是，本文采用了邊緣檢測(cè)的方法進(jìn)行識(shí)別。包括圖像預(yù)處理、邊緣提取、噪聲去除、圖像銳化和領(lǐng)域平均法處理等的主要過(guò)程。最終達(dá)到目標(biāo)識(shí)別的目的。研究結(jié)論如下：
（1）灰度化處理使原來(lái)的彩色圖像灰度化，并進(jìn)行圖像縮放以減少計(jì)算量，提高運(yùn)算效率，對(duì)灰度圖像進(jìn)行邊緣檢測(cè)，濾波和銳化，突出目標(biāo)，削弱背景噪聲。
（2）采用鄰域平均法對(duì)圖像中的目標(biāo)和遠(yuǎn)處水面背景噪聲進(jìn)行分割，提取出圖像中的目標(biāo)，為進(jìn)一步得到目標(biāo)的位置等信息打下基礎(chǔ)。
本文在較廣泛地調(diào)研文獻(xiàn)的基礎(chǔ)上，對(duì)圖像識(shí)別系統(tǒng)進(jìn)行了較為全面的綜述，并以較為大量文字和具體的實(shí)例，通過(guò)使用常用的仿真語(yǔ)言和軟件對(duì)基于數(shù)字圖像處理的障礙物的識(shí)別進(jìn)行了研究，獲得了較理想的識(shí)別結(jié)果。
關(guān)鍵詞： 圖像處理 水面目標(biāo)識(shí)別 邊緣檢測(cè) 視覺(jué)導(dǎo)航
Abstract Digital image processing is the use of computers for image information processing to meet the person's visual psychology or application requirements. Issues discussed at image recognition, is the study of automatically replace people use your computer to work with large amounts of physical information, solve the problem of human beings does not recognize. For the computer, in the actual working environment, image scene has a larger change. Therefore, to distinguish between images fall into which category, often through a series of key technology to achieve. The resulting image recognition also has many kinds of methods.
Based on the target surface obstacles (bamboo, fishing nets, etc.) of image recognition, for example, from the natural light during the day under the condition of image recognition method of study. Focusing on mathematical image processing, this paper discusses on the surface of the water obstacles target image recognition technology and process. Because of the collected image object and background gray color difference is not very obvious, for the image segmentation is not suitable for color segmentation method is used to, so, in this paper, the edge detection method was adopted for identification. Including image preprocessing, edge detection and noise removal and image sharpening and average method, the main process. Ultimately achieve the goal of target recognition. Research conclusions are as follows:
(1)Gray processing make the original color image gray level and image zooming to reduce the amount of calculation, improve the efficiency of operation, and the gray image edge detection, filtering and sharpening, highlight the objectives and weaken the background noise.
(2)The neighborhood average method is adopted to image the target and background noise in the distance the water division, extract the target in the image, the information such as the location of the target is obtained for the further lay the foundation.
In the research literature on the basis of extensively in this paper, the image recognition system was carried out, and with relatively large amounts of text and concrete examples, through the use of the commonly used language and software simulation of obstacle recognition based on digital image processing was studied, and obtained the ideal recognition results.
Key words Image processing Target identification Edge detection Visual navigation
目 錄
第一章 緒 論 1
1.1 研究的背景 1
1.2 本文的研究目的和意義 1
第二章 論文的研究背景 3
2.1 機(jī)器人視覺(jué)導(dǎo)航技術(shù)的應(yīng)用及展望 3
2.2 在工業(yè)檢測(cè)中的應(yīng)用 4
第三章 圖像識(shí)別系統(tǒng) 8
3.1 圖像預(yù)處理 8
3.1.1 灰度化和二值化 9
3.1.2 邊緣檢測(cè) 10
3.2 形態(tài)學(xué)處理 12
3.3 圖像的識(shí)別 14
3.4 小結(jié) 15
第四章 基于數(shù)字圖像處理的障礙物識(shí)別 16
4.1 仿真環(huán)境簡(jiǎn)介 16
4.2 MATLAB處理 18
4.2.1 灰度化和二值化 18
4.2.2 圖像數(shù)據(jù)處理 20
4.2.3 形態(tài)學(xué)處理 21
4.2.4 濾波處理 21
4.2.5 圖像銳化和目標(biāo)識(shí)別 23
4.3 小結(jié) 24
第五章 總結(jié)與展望 25
致 謝 26
參考文獻(xiàn) 27
附錄一（程序代碼） 29
附錄二（程序運(yùn)行結(jié)果） 31
