近十年來,紡織材料圖像處理取得了長足的進(jìn)展。對(duì)混色紗工業(yè)圖像處理的過去、現(xiàn)狀和未來進(jìn)行了調(diào)研,分析了傳統(tǒng)人工視覺和計(jì)算機(jī)視覺用于紗線檢測(cè)的方法及其優(yōu)缺點(diǎn)。接著,探討了大顆粒分析方法。在基于MATLAB的模型基礎(chǔ)上,對(duì)棉結(jié)混雜紗織物的實(shí)時(shí)均勻性進(jìn)行了研究。通過對(duì)現(xiàn)有系統(tǒng)的比較分析發(fā)現(xiàn),大部分混雜紗行業(yè)主要依靠人眼視覺的評(píng)估和經(jīng)驗(yàn)。在MATLAB模型基礎(chǔ)上,提出了一種新的棉結(jié)均勻性檢測(cè)算法,為混紡紗行業(yè)棉結(jié)均勻性的檢測(cè)提供了解決方案。通過棉結(jié)混雜紗的實(shí)時(shí)棉結(jié)檢測(cè)實(shí)驗(yàn),驗(yàn)證了基于核密度函數(shù)和宏像素分析的棉結(jié)檢測(cè)算法,該算法計(jì)算出了91%左右的棉結(jié)不均勻度,這個(gè)結(jié)果無法通過肉眼視覺檢測(cè)達(dá)到。這種新的模型、算法和方法將有助于混雜紗行業(yè)的棉結(jié)均勻性檢測(cè),也可用于其他類型時(shí)尚紗線的檢測(cè)。最后,對(duì)本項(xiàng)研究進(jìn)行了總結(jié),并展望了今后可能的研究方向。 

【文章來源】：東華大學(xué)上海市 211工程院校 教育部直屬院校

【文章頁數(shù)】：64 頁

【學(xué)位級(jí)別】：碩士

【文章目錄】：
摘要
Abstract
Chapter 1 Introduction
    1.1 Background
    1.2 Development in quality assessment of melange yarn
    1.3 Macropixel Analysis to determine homogeneity
    1.4 Significance of this research
    1.5 Creative fruits and thesis outline
Chapter 2 Model of Melange Yarn Inspection with quality controlling
    2.1 Introduction
    2.2 Analysis of visual inspection and its quality controlling
    2.3 Computerized inspection model and system with quality controlling
    2.4 Differences and advantages of the proposed computerized inspectionapproach
    2.5 Summary
Chapter 3 Algorithm & experimental simulations
    3.1 Introduction
    3.2 Algorithm design and workflow analysis of real-time nep detection model
    3.3 Experimental materials and methods
    3.4 Summary
Chapter 4 Result analysis and comparison
    4.1 Introduction
    4.2 Significance of real-time nep detection model
    4.3 Validation of real-time nep detection model
    4.4 Nep detection
    4.5 Homogeneity analysis and comparison
    4.6 Summary
Chapter 5 Conclusion and Future recommendation
    5.1 Conclusion
    5.2 Future recommendation
參考文獻(xiàn) References
附錄 Appendix
    Appendix 1: MATLAB code
攻讀學(xué)位期間發(fā)表的學(xué)術(shù)論文 Published Papers During Postgraduate Years
致謝 Acknowledgement
索引 Index



本文編號(hào)：3244264