ass日本风韵熟妇pics男人扒开女人屁屁桶到爽|扒开胸露出奶头亲吻视频|邻居少妇的诱惑|人人妻在线播放|日日摸夜夜摸狠狠摸婷婷|制服 丝袜 人妻|激情熟妇中文字幕|看黄色欧美特一级|日本av人妻系列|高潮对白av,丰满岳妇乱熟妇之荡,日本丰满熟妇乱又伦,日韩欧美一区二区三区在线

基于氣流脈沖和結構光成像的牛肉嫩度檢測方法
CSTR:
作者:
作者單位:

作者簡介:

通訊作者:

中圖分類號:

基金項目:

國家自然科學基金項目(32071896,、31960487),、江蘇省自然科學基金面上項目(BK20181315)、江蘇省農業(yè)科技自主創(chuàng)新項目(CX(20)3068)和揚州市重點研發(fā)計劃項目(YZ2018038)


Beef Tenderness Detection Based on Pulse Air-puff Combined with Structural Light 3D Imaging
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 圖/表
  • |
  • 訪問統計
  • |
  • 參考文獻
  • |
  • 相似文獻
  • |
  • 引證文獻
  • |
  • 資源附件
  • |
  • 文章評論
    摘要:

    針對傳統牛肉嫩度檢測速度慢,、精度低的問題,,提出了基于氣流脈沖結合結構光3D成像的牛肉嫩度快速無損檢測方法。首先,,利用脈沖氣流對牛肉表面進行沖擊,,同時通過結構光3D成像獲取待測牛肉表面凹陷區(qū)域的三維點云信息;然后,,采用去噪,、點云分割、貪婪投影三角化,、Delaunay三角化,、曲面擬合等算法進行點云處理,獲得牛肉表面凹陷區(qū)域的深度,、映射面積,、表面積和體積等信息;基于此,分別建立基于最小二乘支持向量機回歸(LS-SVR),、BP神經網絡和廣義回歸神經網絡(GRNN)的生鮮牛肉剪切力預測模型,;結果表明,GRNN模型預測表現最佳,,預測集相關系數為0.975,,均方根誤差為5.307N。采用基于K-fold交叉驗證的GRNN神經網絡對牛肉嫩度等級進行預測,結果顯示該方法對較嫩牛肉分級效果較好,,為100%,,對較老牛肉分級效果稍差,為91.3%,。研究表明,,基于氣流脈沖結合結構光3D成像進行牛肉剪切力以及嫩度快速、無損檢測是可行的,。

    Abstract:

    With the aim to solve the problem of low detection speed and precision of beef tenderness, a fast nondestructive detection method for beef tenderness based on airpuff and structural light 3D imaging technology was proposed. The structural light 3D scanning technology was used to obtain the threedimensional point cloud information on the surface of the beef and the point cloud processing technology was combined to extract the depth, area, surface area and volume parameters of the stressed depression area on the beef. In point cloud processing, denoising, point cloud segmentation, greedy projection triangulation, Delaunay triangulation, surface fitting and other algorithms were used to extract the characteristic parameters of beef samples. The modeling method was used to establish the prediction model of beef shear force which about the least squares support vector machine regression (LS-SVR), back propagation (BP) and general regression neural network (GRNN). The results showed that the GRNN model performed the best with the correlation coefficients of prediction set of 0.975, and root mean squared error of 5.307N. The GRNN neural network based on K-fold cross validation was used to predict the tenderness grade. It was worth noting that the results showed that the method had a better grading effect on the tender beef of 100% and a slightly worse grading effect on the tougher beef of 91.3%. The results demonstrated that the proposed airpuff combined with structured light method was effective in beef tenderness detection nondestructively. The research result provided a method for poultry meat tenderness detection and a basis for online poultry tenderness detection which had broad application prospect not only in meat tenderness but also in fruit hardness and ripeness detection.

    參考文獻
    相似文獻
    引證文獻
引用本文

盧偉,胡慶迎,代德建,張澄宇,DENG Yiming.基于氣流脈沖和結構光成像的牛肉嫩度檢測方法[J].農業(yè)機械學報,2020,51(12):324-331. LU Wei, HU Qingying, DAI Dejian, ZHANG Chengyu, DENG Yimin. Beef Tenderness Detection Based on Pulse Air-puff Combined with Structural Light 3D Imaging[J]. Transactions of the Chinese Society for Agricultural Machinery,2020,51(12):324-331.

復制
分享
文章指標
  • 點擊次數:
  • 下載次數:
  • HTML閱讀次數:
  • 引用次數:
歷史
  • 收稿日期:2020-07-26
  • 最后修改日期:
  • 錄用日期:
  • 在線發(fā)布日期: 2020-12-10
  • 出版日期: 2020-12-10
文章二維碼