国产日韩欧美一区二区三区三州_亚洲少妇熟女av_久久久久亚洲av国产精品_波多野结衣网站一区二区_亚洲欧美色片在线91_国产亚洲精品精品国产优播av_日本一区二区三区波多野结衣 _久久国产av不卡

?

基于圖像特征識別的馬鈴薯薯皮粗糙度分級研究

2023-12-29 00:53:54唐振三袁劍龍康亮河程李香呂汰楊晨張峰
中國農(nóng)業(yè)科學(xué) 2023年22期
關(guān)鍵詞:特征參數(shù)粗糙度紋理

唐振三,袁劍龍,康亮河,程李香,呂汰,楊晨,張峰

基于圖像特征識別的馬鈴薯薯皮粗糙度分級研究

唐振三1,袁劍龍1,康亮河2,程李香1,呂汰3,楊晨3,張峰1

1甘肅農(nóng)業(yè)大學(xué)農(nóng)學(xué)院/干旱生境作物學(xué)國家重點實驗室/甘肅省作物遺傳改良與種質(zhì)創(chuàng)新重點實驗室,蘭州 730070;2甘肅農(nóng)業(yè)大學(xué)信息科學(xué)技術(shù)學(xué)院,蘭州 730070;3天水市農(nóng)業(yè)科學(xué)研究所,天水 741000

【目的】馬鈴薯薯皮粗糙度分級研究可以提供塊莖外觀品質(zhì)性狀無損檢測方法,為客觀評價品質(zhì)質(zhì)量和高通量篩選品種提供理論和實踐基礎(chǔ)。【方法】以79份馬鈴薯品種(系)為供試材料,利用相機采集有/無芽眼的薯皮圖像?;贛ATLAB R2016a軟件對薯皮圖像預(yù)處理,隨機選擇8份材料用相關(guān)函數(shù)指標(biāo)比較圖像灰度化、增強及去噪效果。利用灰度共生矩陣(gray level co-occurrence matrix,GLCM)提取圖像特征參數(shù)角二階矩(angular second moment,ASM)、熵(entropy,ENT)、對比度(contrast,CON)和相關(guān)度(correlation,COR),并確定矩陣最適像素距離(d)。比較兩類薯皮圖像特征參數(shù)間的差異,選擇差異較小的薯皮圖像特征集進(jìn)行統(tǒng)計分析和分類識別。構(gòu)建支持向量機(support vector machines,SVM)和BP神經(jīng)網(wǎng)絡(luò)(backpropagation neural network,BPNN)模型對薯皮粗糙度分級分類,模型分級精度評價指標(biāo)為準(zhǔn)確率、精準(zhǔn)率、召回率及調(diào)和平均數(shù)?!窘Y(jié)果】加權(quán)平均值法進(jìn)行灰度處理后的薯皮圖像紋理結(jié)構(gòu)清晰,圖像清晰度評價值為2.5698±0.5959,顯著高于平均值法(1.8035±0.4856)和最大值法(1.0535±0.4088);直方圖均衡化增強后的薯皮圖像灰度級范圍由100—200擴大為0—200,灰度分布更加廣泛;中值濾波對3×3窗口下的薯皮圖像椒鹽噪聲去噪效果明顯,峰值信噪比(peak signal-to-noise ratio,PSNR)最大((28.6250±3.9784)Bp),顯著高于3×3和5×5窗口下對高斯噪聲去噪后的PSNR。通過GLCM(d=4)提取的兩類薯皮圖像特征參數(shù)間差異顯著,選擇其中差異較小的無芽眼薯皮圖像特征集進(jìn)行統(tǒng)計分析和分類識別,結(jié)果表明該特征集變異系數(shù)差異明顯,對比度變異系數(shù)最大(0.40),其次是角二階矩(0.24)和相關(guān)度(0.23),熵變異系數(shù)最小(0.18)。將該特征集作為分類模型輸入變量用于薯皮分類,相較于BP神經(jīng)網(wǎng)絡(luò),SVM對馬鈴薯薯皮粗糙度的整體分類性能較高,準(zhǔn)確率為87.5%。其中,對光滑皮和重麻皮的預(yù)測準(zhǔn)確度和識別能力最高,精準(zhǔn)率均為100%,召回率分別為85.7%和100%,調(diào)和平均數(shù)分別為92.3%和100%?!窘Y(jié)論】綜合利用本研究提出的圖像處理技術(shù)及GLCM提取的紋理特征參數(shù)能有效表征馬鈴薯塊莖薯皮粗糙度差異;通過構(gòu)建SVM分類模型可實現(xiàn)基于機器視覺的馬鈴薯薯皮粗糙度分級,且準(zhǔn)確率達(dá)87.5%。

馬鈴薯;薯皮粗糙度;圖像特征;機器視覺;支持向量機

0 引言

【研究意義】馬鈴薯是世界第四大糧食作物。薯皮粗糙度是馬鈴薯塊莖重要外觀品質(zhì)性狀,是品種特異性測試及商業(yè)分類分級的重要標(biāo)準(zhǔn)和依據(jù)[1]?!厩叭搜芯窟M(jìn)展】薯皮粗糙度由遺傳因子和環(huán)境因子相互作用決定,是細(xì)胞生長、分裂、分化和代謝相互作用的最終體現(xiàn)[2]。目前馬鈴薯分級篩選主要側(cè)重塊莖形狀、大小及缺陷的簡單人工分級和機械分級,人工分級依賴視覺感官定性分類,通量和精準(zhǔn)度低;機械分級雖可避免人工差異,但會造成不同程度的二次損傷,降低馬鈴薯商品性[3-4]。新興的基于圖像處理的機器視覺技術(shù)可通過對圖像顏色、紋理、形狀等特征信息的挖掘,快速實現(xiàn)待測物精準(zhǔn)高效無損鑒別[4-5]。圖像處理中,紋理粗糙度是圖像不依賴顏色和亮度變化的視覺特征,表現(xiàn)出局部不規(guī)則、宏觀有規(guī)律的特性,可以反映圖像像素間的空間分布關(guān)系[6]。利用灰度共生矩陣(gray level co-occurrence matrix,GLCM)提取的紋理特征參數(shù)角二階矩(angular second moment,ASM)、熵(entropy,ENT)、對比度(contrast,CON)和相關(guān)度(correlation,COR)等,通過與機器學(xué)習(xí)算法結(jié)合,建立BP神經(jīng)網(wǎng)絡(luò)、支持向量機(support vector machines,SVM)等分類模型可快速實現(xiàn)農(nóng)產(chǎn)品無損傷檢測[7-15],已被廣泛應(yīng)用于玉米籽粒識別分類、病蟲害識別、種子質(zhì)量檢測及水果自動分揀系統(tǒng)構(gòu)建等領(lǐng)域[12,16-19]?!颈狙芯壳腥朦c】目前,基于圖像處理的機器視覺技術(shù)可在馬鈴薯生產(chǎn)加工過程中實現(xiàn)精準(zhǔn)高效的質(zhì)量控制和定量鑒別。但研究方向主要集中在馬鈴薯塊莖形狀大小、表皮缺陷及病害鑒別等方面[3-5,16,20-21],薯皮粗糙度分級應(yīng)用中尚無涉及?!緮M解決的關(guān)鍵問題】擬通過基于圖像處理的機器視覺技術(shù)對馬鈴薯薯皮粗糙度分級展開研究,找出合適的薯皮圖像處理方法及特征參數(shù),構(gòu)建高識別率分類模型用于薯皮粗糙度分類,為馬鈴薯薯皮粗糙度分級識別提供高效精準(zhǔn)且系統(tǒng)的客觀評價方法,加快馬鈴薯自動化生產(chǎn)和商品化處理效率。

1 材料與方法

1.1 試驗材料

以79份馬鈴薯品種(系)為供試材料(表1),2021年種植于甘肅省渭源縣五竹鎮(zhèn)鹿鳴村(海拔2 260 m,年平均降雨量520—560 mm,年平均氣溫5.8 ℃,年平均日照時數(shù)約2 412 h,無霜期145 d,黑壚土)。

1.2 薯皮圖像采集及人工分級

使用專業(yè)可調(diào)LED光源拍攝箱(60 cm×60 cm×60 cm)拍照,SMART SENSOR光照度計(測量范圍1—200 000 lx,分辨率1 lx)測定光照強度為(9 000± 50)lx。選擇相機型號Canon EOS 760 D,鏡頭EF-S 18-135 mm f/3.5—5.6 IS STM。相機參數(shù)設(shè)定:快門速度1/100 s,光圈值(F)6.3,焦距50 mm,感光度(ISO)100,相機固定于拍攝物體上方,物距60 cm,圖像像素6 000×4 000。各塊莖圖像選取6個像素大小為1 000×1 000的區(qū)域作為參試樣本圖像。

參照馬鈴薯品種特異性、一致性和穩(wěn)定性測試指南(GB/T 19557.28—2018),將采集的馬鈴薯人工分級[1],用作分類模型結(jié)果對照。

1.3 研究方法

用MATLAB R2016a對薯皮圖像預(yù)處理(灰度化、增強和降噪);將處理后的圖像劃分為有、無芽眼兩類,芽眼圖像按1 000×1 000像素區(qū)域分塊,再根據(jù)有無芽眼對每個區(qū)域進(jìn)行分類。利用GLCM提取兩類薯皮圖像紋理粗糙度特征并比較差異,選擇對分類識別影響較小的圖像作為SVM和BP神經(jīng)網(wǎng)絡(luò)分類模型輸入變量用于薯皮粗糙度分類識別,采用混淆矩陣指標(biāo)(準(zhǔn)確率、精準(zhǔn)率、召回率和F1值)評價分類精度(圖1)。

圖1 技術(shù)路線

1.3.1 薯皮圖像預(yù)處理 采用最大值法、平均值法和加權(quán)平均值法進(jìn)行圖像灰度處理[22],從中隨機選取8份材料通過能量梯度函數(shù)比較三者處理效果[23-24];采用直方圖均衡化法進(jìn)行圖像灰度增強[25];采用中值濾波在3×3和5×5窗口下對添加高斯噪聲和椒鹽噪聲的薯皮圖像降噪,計算峰值信噪比(peak signal-to- noise ratio,PSNR)評價預(yù)處理后的圖像質(zhì)量,PSNR 值越高,越接近原圖[26-28]。

1.3.2 薯皮圖像分類處理 根據(jù)有無芽眼特征對預(yù)處理薯皮圖像進(jìn)行區(qū)分,對含芽眼的薯皮圖像按1 000×1 000像素大小分塊,再將各區(qū)域按有無芽眼進(jìn)行分類。

1.3.3 薯皮圖像紋理特征參數(shù)提取及差異比較 將待處理圖像灰度保持256級,選取圖像0°、45°、90°和135°方向均值作為GLCM方向參數(shù)(q)[14,29-30];利用GLCM計算選取的8份參試材料特征參數(shù)在不同像素距離處的取值以確定矩陣最適像素距離(d),然后通過GLCM提取有、無芽眼的薯皮圖像特征參數(shù)角二階距、對比度、相關(guān)度和熵進(jìn)行差異比較。公式詳見(1)—(4)[22,30-33]。

角二階距:GLCM元素值平方和,反映圖像灰度分布均勻程度和紋理粗細(xì)程度。

對比度:反映圖像清晰度和紋理溝紋深淺程度。

相關(guān)度:度量GLCM元素在行或列方向上的相似程度,反映圖像局部灰度相關(guān)性。

熵:圖像所含信息量的度量,反映圖像紋理非均勻程度或復(fù)雜程度。

1.3.4 薯皮粗糙度分類識別 采用SVM和BP神經(jīng)網(wǎng)絡(luò)對薯皮粗糙度分類評價,在特征參數(shù)中隨機選取55份數(shù)據(jù)用作訓(xùn)練,剩余24份數(shù)據(jù)用作測試驗證。在SVM中,設(shè)定輸入?yún)?shù)和輸出參數(shù)為4,徑向基函數(shù)作為核函數(shù),懲罰系數(shù)ζ和非負(fù)松弛項g分別設(shè)定為25和35;設(shè)定BP神經(jīng)網(wǎng)絡(luò)輸入層、輸出層和隱含層節(jié)點數(shù)分別為4、1、6,最大迭代數(shù)為1 000。相關(guān)公式見文獻(xiàn)[34]。

1.3.5 數(shù)據(jù)處理及分類模型精度評價 用Microsoft Excel 2010和SPSS 22.0進(jìn)行薯皮圖像粗糙度特征數(shù)據(jù)的統(tǒng)計描述和方差分析;用準(zhǔn)確率(accuracy,Acc)、精準(zhǔn)率(precision,Pre)、召回率(recall,Re)和調(diào)和平均數(shù)(F1)評價模型分類結(jié)果。公式詳見(5)—(8)[35]:

Acc=(TP+TN)/(TP+TN+FP+FN) (5)

Pre=TP/(TP+FP) (6)

Re=TP/(TP+FN) (7)

F1=2Pre×Re/(Pre+Re) (8)

式中,TN、FN、TP和FP分別為真反例、假反例、真正例和假正例。

2 結(jié)果

2.1 薯皮圖像人工分級

參照馬鈴薯品種特異性、一致性和穩(wěn)定性測試指南(GB/T 19557.28—2018),將79份馬鈴薯品種(系)人工分為重麻皮(6份)、麻皮(26份)、略麻皮(27份)和光滑皮(20份),列出分級結(jié)果的部分馬鈴薯薯皮圖像(表2)。

2.2 圖像預(yù)處理結(jié)果分析

由表3可知,加權(quán)平均值法處理后的圖像紋理結(jié)構(gòu)最清晰,效果最好。平均值法、最大值法和加權(quán)平均值法處理的圖像灰度清晰度差異顯著,加權(quán)平均值法與平均值法和最大值法的均值差最大,分別為0.7663和1.1516(表4)。

薯皮圖像均衡化增強后,灰度級范圍由100—200(表5,C-1、C-2、C-3和C-4)擴大為0—200,灰度分布更為均勻廣泛(表5,D-1、D-2、D-3和D-4)。相較于表5原始圖像A-1、A-2、A-3和A-4,增強后的薯皮圖像B-1、B-2、B-3和B-4紋理更加清晰突出。

利用中值濾波對圖像去噪,發(fā)現(xiàn)中值濾波對含椒鹽噪聲的圖像去噪效果顯著,在3×3窗口下中值濾波濾除椒鹽噪聲的PSNR均值差最大,且在3×3窗口下中值濾波對椒鹽噪聲濾除效果最好,最接近原始圖像(表6和表7)。

2.3 GLCM構(gòu)造因子確定

像素距離d=4時,GLCM提取的8個參試材料薯皮圖像特征參數(shù)變化基本趨于一致,表現(xiàn)出穩(wěn)定性(圖2)。

2.4 特征參數(shù)統(tǒng)計分析

2.4.1 薯皮圖像特征參數(shù)差異比較 利用GLCM提取有、無芽眼的薯皮圖像特征參數(shù)進(jìn)行差異比較。相較于無芽眼的薯皮圖像,芽眼薯皮圖像特征參數(shù)除角二階矩外,對比度、熵和相關(guān)度均存在極顯著差異(圖3)。

2.4.2 特征參數(shù)變異系數(shù) 為避免芽眼對薯皮紋理特征的識別影響,選擇差異較小的無芽眼薯皮圖像進(jìn)行分類識別。經(jīng)統(tǒng)計分析,薯皮圖像紋理特征參數(shù)變異系數(shù)差異明顯,對比度變異系數(shù)最大(0.40),其次是角二階矩(0.24)和相關(guān)度(0.23),熵變異系數(shù)最小(0.18)(表8)。

表2 薯皮圖像

A:重麻皮;B:麻皮;C:略麻皮;D:光滑皮

A: Heavy hemp skin; B: Hemp skin; C: Slightly hemp skin; D: Smooth skin

表3 灰度方法處理的馬鈴薯薯皮圖像效果比較

A:原始圖像;B:平均值法;C:最大值法;D:加權(quán)平均值法

A: Original images B: Average method; C: Maximum method; D: Weighted-average method

表4 灰度圖像清晰度差異比較

A:平均值法;B:最大值法;C:加權(quán)平均值法。不同小寫字母表示多重比較差異顯著(<0.05)

A: Average method; B: Maximum method; C: Weighted-average method. Different lowercase letters indicate significant differences in multiple comparisons (<0.05)

A:原始圖像;B:直方圖均衡化處理后圖像;C:處理前圖像灰度直方圖;D:處理后圖像灰度直方圖

A: Original images; B: Images after histogram equalization; C: Gray scale histogram of images before processing; D: Gray scale histogram of images after processing

2.5 薯皮粗糙度分類評價

分類模型中,相較于BP神經(jīng)網(wǎng)絡(luò),SVM在薯皮分類中準(zhǔn)確率較高(87.5%),表明該模型對薯皮類別整體分類性能較好,且對光滑皮和重麻皮類別的預(yù)測準(zhǔn)確度和識別能力最高,精準(zhǔn)率均為100%,召回率分別為85.7%和100%,F(xiàn)1值分別為92.3%和100%(表9)。

3 討論

3.1 相機拍攝設(shè)置影響圖像質(zhì)量

薯皮圖像采集過程中,受環(huán)境條件、攝像設(shè)備等因素影響,圖像質(zhì)量可能存在差異。像素距離大小主要取決于拍攝圖像尺寸、感光度(ISO)、鏡頭光圈等參數(shù)的設(shè)置[36-37]。像素距離太小,放大尺寸的圖像會出現(xiàn)模糊現(xiàn)象;低光環(huán)境下拍攝,為提高相機鏡頭感光性能,應(yīng)選較大像素距離。

表6 圖像降噪處理比較

A:重麻皮;B:麻皮;C:略麻皮;D:光滑皮 A: Heavy hemp skin; B: Hemp skin; C: Slightly hemp skin; D: Smooth skin

A:G16;B:大西洋;C:布爾班克;D:G67;E:夏波蒂;F:G121;G:隴薯10號;H:G33 A: G16; B: Atlantic; C: Burbank; D: G67; E: Shepody; F: G121; G: Longshu no.10; H: G33

表7 去噪圖像峰值信噪比(PSNR)差異比較

A:3×3椒鹽噪聲;B:5×5椒鹽噪聲;C:3×3高斯噪聲;D:5×5高斯噪聲。不同小寫字母表示多重比較差異顯著(<0.05)

A: The 3×3 salt and pepper noise; B: The 5×5 salt and pepper noise; C: The 3×3 gaussian noise; D: The 5×5 gaussian noise. Different lowercase letters indicate significant differences in multiple comparisons (<0.05)

表8 灰度共生矩陣提取薯皮特征參數(shù)表現(xiàn)及變異系數(shù)

表9 分類精度結(jié)果

3.2 圖像處理方法的合理選擇

合理的圖像處理是圖像特征準(zhǔn)確識別的前提。圖像灰度化是圖像處理的基本方法,選擇合適灰度化方法可提高圖像灰度效果控制[38]。本研究采用的加權(quán)平均值法考慮了R(Red)、G(Green)、B(Blue)三通道權(quán)值分配,處理后的圖像相較于平均值法和最大值法,紋理清晰度最高,更符合人眼視覺感受。由于圖像灰度分布在較窄區(qū)間,亮度過于集中,使薯皮紋理模糊。采用直方圖均衡化在0—255灰度級范圍內(nèi)調(diào)整圖像灰度直方圖分布,拓展像素集中區(qū)域,歸并低頻像素,可使圖像灰度分布均勻且動態(tài)范圍擴大,圖像明暗對比明顯。但對低頻灰度級過度簡并,可能會造成圖像邊緣細(xì)節(jié)信息丟失。圖像灰度集中區(qū)域存在直方圖高峰,均衡化后灰度級過度拉伸會導(dǎo)致圖像對比度過度增強和出現(xiàn)偽影[39-40]。去噪可規(guī)避圖像處理中噪聲干擾造成的圖像信息丟失并保護(hù)圖像邊緣細(xì)節(jié)[26]。在實際應(yīng)用中,忽略噪聲類型盲目采用濾波算法去噪難以達(dá)到預(yù)期效果。本研究采用的中值濾波是一種非線性信號處理方法,對椒鹽噪聲、脈沖噪聲有明顯的抑制作用。為驗證中值濾波的去噪效果,利用PSNR客觀衡量去噪后的圖像質(zhì)量,為選擇合適的去噪方法提供科學(xué)依據(jù),以實現(xiàn)圖像有效去噪[41-42]。

A:角二階距;B:對比度;C:熵;D:相關(guān)度。ns:不顯著;**:差異極顯著(P<0.01)

3.3 薯皮紋理特征及有限樣本量對機器學(xué)習(xí)方法分類識別的影響

薯皮圖像紋理特征識別中,紋理是不依賴于顏色、亮度變化而反映圖像同質(zhì)現(xiàn)象的視覺特征[12]。利用共生矩陣提取紋理特征用于圖像相似性度量檢索時,在0°、45°、90°和135°四個方向上求取特征均值且歸一化,可使圖像發(fā)生旋轉(zhuǎn)和縮放時對共生矩陣提取的紋理特征影響非常小,且對視覺變化、仿射變化及噪聲干擾起到穩(wěn)定作用[34,43]。薯皮凸起、凹陷、裂變等缺陷會對薯皮紋理分布和連續(xù)性產(chǎn)生擾動,使紋理分布不均勻。薯皮缺陷常與周圍區(qū)域形成明顯對比,紋理特征存在差異,在紋理識別中,可能會對SVM和BP神經(jīng)網(wǎng)絡(luò)模型分類性能產(chǎn)生影響,增大識別難度。模型應(yīng)用中應(yīng)針對薯皮缺陷調(diào)整訓(xùn)練策略,選擇適當(dāng)?shù)腟VM核函數(shù)和懲罰系數(shù),或者在BP神經(jīng)網(wǎng)絡(luò)中利用交叉驗證方法選擇最佳隱藏層節(jié)點數(shù)和層數(shù),以此提高模型學(xué)習(xí)能力,以便更好地識別圖像紋理特征[44]。

利用分類模型進(jìn)行薯皮分類,有限的樣本量會使模型過擬合,導(dǎo)致分類結(jié)果不穩(wěn)定、泛化性能低,無法在實際場景中應(yīng)用。選擇合適的分類模型是有效降低樣本量過小導(dǎo)致模型過擬合和泛化性能低的重要方法[3,15,38]。本研究采用SVM和BP神經(jīng)網(wǎng)絡(luò)模型對不同薯皮類別的馬鈴薯分類,SVM模型對薯皮的整體分類性能顯著高于BP神經(jīng)網(wǎng)絡(luò),更適于解決小樣本非線性可分問題。其可避免過度依賴樣本數(shù)據(jù),降低模型對訓(xùn)練集的過擬合,提高對新樣本數(shù)據(jù)的分類準(zhǔn)確度。另外,為避免樣本量影響分類模型的可靠性和實用性,通過設(shè)定薯皮圖像像素大小,從單一樣本中獲取多幅圖像擴充樣本數(shù)據(jù)集,可使分類模型魯棒性和泛化性能達(dá)到平衡。

4 結(jié)論

加權(quán)平均值法、直方圖均衡化法和中值濾波法能夠?qū)崿F(xiàn)圖像灰度化、增強及去噪效果;灰度共生矩陣提取的圖像紋理特征參數(shù)角二階矩、對比度、熵和相關(guān)度作為分類模型輸入值用作馬鈴薯薯皮識別,可有效表征馬鈴薯薯皮粗糙度差異。小樣本集中,支持向量機模型識別性能優(yōu)于BP神經(jīng)網(wǎng)絡(luò),分類準(zhǔn)確率達(dá)87.5%,更適合薯皮粗糙度分類。

[1] 國家市場監(jiān)督管理總局, 中國國家標(biāo)準(zhǔn)化管理委員會. 植物品種特異性、一致性和穩(wěn)定性測試指南馬鈴薯: GB/T 19557.28—2018. 北京: 中國標(biāo)準(zhǔn)出版社, 2018.

State Administration for Market Regulation; Standardization Administration of the People's Republic of China. Guidelines for the conduct of tests for distinctness, uniformity and stability-Potato (L.): GB/T 19557.28-2018. Beijing: Standard Press of China, 2018. (in Chinese)

[2] 韓仲志, 趙友剛. 利用花生莢果圖像特征識別品種與檢驗種子. 作物學(xué)報, 2012, 38(3): 535-540.

HAN Z Z, ZHAO Y G. Variety identification and seed test by peanut pod image characteristics. Acta Agronomica Sinica, 2012, 38(3): 535-540. (in Chinese)

[3] 王紅軍, 熊俊濤, 黎鄒鄒, 鄧建猛, 鄒湘軍. 基于機器視覺圖像特征參數(shù)的馬鈴薯質(zhì)量和形狀分級方法. 農(nóng)業(yè)工程學(xué)報, 2016, 32(8): 272-277.

WANG H J, XIONG J T, LI Z Z, DENG J M, ZOU X J. Potato grading method of weight and shape based on imaging characteristics parameters in machine vision system. Transactions of the Chinese Society of Agricultural Engineering, 2016, 32(8): 272-277. (in Chinese)

[4] 姜宏, 于永波, 章翔峰, 陳宇彤. 馬鈴薯外部品質(zhì)分級方法綜述. 科學(xué)技術(shù)與工程, 2022, 22(14): 5519-5527.

JIANG H, YU Y B, ZHANG X F, CHEN Y T. Potato external quality grading methods: a review. Science Technology and Engineering, 2022, 22(14): 5519-5527. (in Chinese)

[5] 周竹, 黃懿, 李小昱, 文東東, 汪成龍, 陶海龍. 基于機器視覺的馬鈴薯自動分級方法. 農(nóng)業(yè)工程學(xué)報, 2012, 28(7): 178-183.

ZHOU Z, HUANG Y, LI X Y, WEN D D, WANG C L, TAO H L. Automatic detecting and grading method of potatoes based on machine vision. Transactions of the Chinese Society of Agricultural Engineering, 2012, 28(7): 178-183. (in Chinese)

[6] 劉麗, 匡綱要. 圖像紋理特征提取方法綜述. 中國圖象圖形學(xué)報, 2009, 14(4): 622-635.

LIU L, KUANG G Y. Overview of image textural feature extraction methods. Journal of Image and Graphics, 2009, 14(4): 622-635. (in Chinese)

[7] 展慧, 李小昱, 王為, 汪成龍, 周竹, 黃懿. 基于機器視覺的板栗分級檢測方法. 農(nóng)業(yè)工程學(xué)報, 2010, 26(4): 327-331.

ZHAN H, LI X Y, WANG W, WANG C L, ZHOU Z, HUANG Y. Determination of chestnuts grading based on machine vision. Transactions of the Chinese Society of Agricultural Engineering, 2010, 26(4): 327-331. (in Chinese)

[8] 薄華, 馬縛龍, 焦李成. 圖像紋理的灰度共生矩陣計算問題的分析. 電子學(xué)報, 2006, 34(1): 155-158, 134.

BO H, MA F L, JIAO L C. Research on computation of GLCM of image texture. Acta Electronica Sinica, 2006, 34(1): 155-158, 134. (in Chinese)

[9] SINGH K R, CHAUDHURY S. Comparative analysis of texture feature extraction techniques for rice grain classification. IET Image Processing, 2020, 14(11): 2532-2540.

[10] LADDI A, SHARMA S, KUMAR A, KAPUR P. Classification of tea grains based upon image texture feature analysis under different illumination conditions. Journal of Food Engineering, 2013, 115(2): 226-231.

[11] SHAFI U, MUMTAZ R, HAQ I U, HAFEEZ M, IQBAL N, SHAUKAT A, ZAIDI S M H, MAHMOOD Z. Wheat yellow rust disease infection type classification using texture features. Sensors, 2021, 22(1): 146.

[12] 張超, 喬敏, 劉哲, 金虹杉, 寧明宇, 孫海艷. 基于無人機和衛(wèi)星遙感影像的制種玉米田識別紋理特征尺度優(yōu)選. 農(nóng)業(yè)工程學(xué)報, 2017, 33(17): 98-104.

ZHANG C, QIAO M, LIU Z, JIN H S, NING M Y, SUN H Y. Texture scale analysis and identification of seed maize fields based on UAV and satellite remote sensing images. Transactions of the Chinese Society of Agricultural Engineering, 2017, 33(17): 98-104. (in Chinese)

[13] 謝元澄, 徐煥良, 謝莊. 基于牛肉大理石花紋標(biāo)準(zhǔn)(BMS)圖像的紋理特征分析. 中國農(nóng)業(yè)科學(xué), 2010, 43(24): 5121-5128.doi: 10.3864/j.issn.0578-1752.2010.24.016.

XIE Y C, XU H L, XIE Z. Analysis of texture features based on beef marbling standards (BMS) images. Scientia Agricultura Sinica, 2010, 43(24): 5121-5128. doi: 10.3864/j.issn.0578-1752.2010.24.016. (in Chinese)

[14] 高程程, 惠曉威. 基于灰度共生矩陣的紋理特征提取. 計算機系統(tǒng)應(yīng)用, 2010, 19(6): 195-198.

GAO C C, HUI X W. GLCM-based texture feature extraction. Computer Systems & Applications, 2010, 19(6): 195-198. (in Chinese)

[15] 劉濤, 仲曉春, 孫成明, 郭文善, 陳瑛瑛, 孫娟. 基于計算機視覺的水稻葉部病害識別研究. 中國農(nóng)業(yè)科學(xué), 2014, 47(4): 664-674.doi: 10.3864/j.issn.0578-1752.2014.04.006.

LIU T, ZHONG X C, SUN C M, GUO W S, CHEN Y Y, SUN J. Recognition of rice leaf diseases based on computer vision. Scientia Agricultura Sinica, 2014, 47(4): 664-674. doi: 10.3864/j.issn.0578- 1752.2014.04.006. (in Chinese)

[16] 黨滿意, 孟慶魁, 谷芳, 顧彪, 胡耀華. 基于機器視覺的馬鈴薯晚疫病快速識別. 農(nóng)業(yè)工程學(xué)報, 2020, 36(2): 193-200.

DANG M Y, MENG Q K, GU F, GU B, HU Y H. Rapid recognition of potato late blight based on machine vision. Transactions of the Chinese Society of Agricultural Engineering, 2020, 36(2): 193-200. (in Chinese)

[17] GOMES J F S, LETA F R. Applications of computer vision techniques in the agriculture and food industry: A review. European Food Research and Technology, 2012, 235(6): 989-1000.

[18] GUZMáN E, BAETEN V, PIERNA J A F, GARCíA-MESA J A. Determination of the olive maturity index of intact fruits using image analysis. Journal of Food Science and Technology, 2015, 52(3): 1462-1470.

[19] WANG A C, ZHANG W, WEI X H. A review on weed detection using ground-based machine vision and image processing techniques. Computers and Electronics in Agriculture, 2019, 158(2019): 226-240.

[20] LI B, XU X M, ZHANG L, HAN J W, BIAN C S, LI G C, LIU J G, JIN L P. Above-ground biomass estimation and yield prediction in potato by using UAV-based RGB and hyperspectral imaging. ISPRS Journal of Photogrammetry and Remote Sensing, 2020, 162: 161-172.

[21] RAZMJOOY N, MOUSAVI B S, SOLEYMANI F. A real-time mathematical computer method for potato inspection using machine vision. Computers & Mathematics with Applications, 2012, 63(1): 268-279.

[22] 紀(jì)娜, 何國榮. 基于灰度識別的獼猴桃形狀疤痕圖像紋理特征提取方法. 自動化與儀器儀表, 2019(1): 159-162.

JI N, HE G R. Texture feature extraction method of kiwifruit shape scar image based on gray level recognition. Automation & Instrumentation, 2019(1): 159-162. (in Chinese)

[23] 李祚林, 李曉輝, 馬靈玲, 胡玥, 唐伶俐. 面向無參考圖像的清晰度評價方法研究. 遙感技術(shù)與應(yīng)用, 2011, 26(2): 239-246.

LI Z L, LI X H, MA L L, HU Y, TANG L L. Research of definition assessment based on no-reference digital image quality. Remote Sensing Technology and Application, 2011, 26(2): 239-246. (in Chinese)

[24] 宋鳳菲. 彩色圖像灰度化及其效果的客觀評價方法研究[D]. 泉州: 華僑大學(xué), 2014.

SONG F F. Research on objective evaluation method of color image graying and its effect [D]. Quanzhou: Huaqiao University, 2014. (in Chinese)

[25] 楊衛(wèi)中, 徐銀麗, 喬曦, 饒偉, 李道亮, 李振波. 基于對比度受限直方圖均衡化的水下海參圖像增強方法. 農(nóng)業(yè)工程學(xué)報, 2016, 32(6): 197-203.

YANG W Z, XU Y L, QIAO X, RAO W, LI D L, LI Z B. Method for image intensification of underwater sea cucumber based on contrast-limited adaptive histogram equalization. Transactions of the Chinese Society of Agricultural Engineering, 2016, 32(6): 197-203. (in Chinese)

[26] 孫宏琦, 施維穎, 巨永鋒. 利用中值濾波進(jìn)行圖像處理. 長安大學(xué)學(xué)報(自然科學(xué)版), 2003, 23(2): 104-106.

SUN H Q, SHI W Y, JU Y F. Image processing with medium value filter. Journal of Chang’an University (Natural Science Edition), 2003, 23(2): 104-106. (in Chinese)

[27] 劉祝華. 圖像去噪方法的研究[D]. 南昌: 江西師范大學(xué), 2005.

LIU Z H. Research on image denoising method [D]. Nanchang: Jiangxi Normal University, 2005. (in Chinese)

[28] DASS A K. Improvising MSN and PSNR for finger-print image noised by GAUSSIAN and SALT & PEPPER. The International Journal of Multimedia & Its Applications, 2012, 4(4): 59-72.

[29] HARALICK R M, SHANMUGAM K, DINSTEIN I. Textural features for image classification. IEEE Transactions on Systems, Man, and Cybernetics, 1973, SMC-3(6): 610-621.

[30] 侯群群, 王飛, 嚴(yán)麗. 基于灰度共生矩陣的彩色遙感圖像紋理特征提取. 自然資源遙感, 2013, 25(4): 26-32.

HOU Q Q, WANG F, YAN L. Extraction of color image texture feature based on gray-level co-occurrence matrix. Remote Sensing for Natural Resources, 2013, 25(4): 26-32. (in Chinese)

[31] 蔡葦荻, 張羽, 劉海燕, 鄭恒彪, 程濤, 田永超, 朱艷, 曹衛(wèi)星, 姚霞. 基于成像高光譜的小麥冠層白粉病早期監(jiān)測方法. 中國農(nóng)業(yè)科學(xué), 2022, 55(6): 1110-1126.doi: 10.3864/j.issn.0578-1752.2022. 06.005.

CAI W D, ZHANG Y, LIU H Y, ZHENG H B, CHENG T, TIAN Y C, ZHU Y, CAO W X, YAO X. Early detection on wheat canopy powdery mildew with hyperspectral imaging. Scientia Agricultura Sinica, 2022, 55(6): 1110-1126. doi: 10.3864/j.issn.0578-1752.2022. 06.005. (in Chinese)

[32] 李智峰, 朱谷昌, 董泰鋒. 基于灰度共生矩陣的圖像紋理特征地物分類應(yīng)用. 地質(zhì)與勘探, 2011, 47(3): 456-461.

LI Z F, ZHU G C, DONG T F. Application of GLCM-based texture features to remote sensing image classification. Geology and Exploration, 2011, 47(3): 456-461. (in Chinese)

[33] 鄭冠楠, 譚豫之, 張俊雄, 李偉. 基于計算機視覺的馬鈴薯自動檢測分級. 農(nóng)業(yè)機械學(xué)報, 2009, 40(4): 166-168, 156.

ZHENG G N, TAN Y Z, ZHANG J X, LI W. Automatic detecting and grading method of potatoes with computer vision. Transactions of the Chinese Society for Agricultural Machinery, 2009, 40(4): 166-168, 156. (in Chinese)

[34] 黃辰, 費繼友. 基于圖像特征融合的蘋果在線分級方法. 農(nóng)業(yè)工程學(xué)報, 2017, 33(1): 285-291.

HUANG C, FEI J Y. Online apple grading based on decision fusion of image features. Transactions of the Chinese Society of Agricultural Engineering, 2017, 33(1): 285-291. (in Chinese)

[35] 魏永康, 楊天聰, 臧少龍, 賀利, 段劍釗, 謝迎新, 王晨陽, 馮偉. 基于無人機多光譜影像特征融合的小麥倒伏監(jiān)測. 中國農(nóng)業(yè)科學(xué), 2023, 56(9): 1670-1685.doi: 10.3864/j.issn.0578-1752.2023.09.005.

WEI Y K, YANG T C, ZANG S L, HE L, DUAN J Z, XIE Y X, WANG C Y, FENG W. Monitoring wheat lodging based on UAV multi-spectral image feature fusion. Scientia Agricultura Sinica, 2023, 56(9): 1670-1685. doi: 10.3864/j.issn.0578-1752.2023.09.005. (in Chinese)

[36] 高傳新, 殷勇. 數(shù)碼相機ISO感光度的調(diào)整對刑事案件中痕跡檢驗照相質(zhì)量的影響. 影像技術(shù), 2008, 21(3): 33-35.

GAO C X, YIN Y. Influence of digital camera ISO speed setting on the quality of criminological photography for traces detection. Image Technology, 2008, 21(3): 33-35. (in Chinese)

[37] PHAM L H, TRAN D N N, RHIE C H, JEON J W. Analysis of the smartphone camera exposure effect on laser extraction//2021 International Conference on Electronics, Information, and Communication (ICEIC), 2021.

[38] 唐俊, 鄧立苗, 陳輝, 欒濤, 馬文杰. 基于機器視覺的玉米葉片透射圖像特征識別研究. 中國農(nóng)業(yè)科學(xué), 2014, 47(3): 431-440. doi: 10.3864/j.issn.0578-1752.2014.03.003.

TANG J, DENG L M, CHEN H, LUAN T, MA W J. Research on maize leaf recognition of characteristics from transmission image based on machine vision. Scientia Agricultura Sinica, 2014, 47(3): 431-440. doi: 10.3864/j.issn.0578-1752.2014.03.003. (in Chinese)

[39] 楊嘉能. 基于直方圖均衡的圖像增強算法優(yōu)化研究[D]. 烏魯木齊: 新疆大學(xué), 2021.

YANG J N. Optimization of image enhancement algorithm based on histogram equalization [D]. Urumqi: Xinjiang University, 2021. (in Chinese)

[40] 陳永亮. 灰度圖像的直方圖均衡化處理研究[D]. 合肥: 安徽大學(xué), 2014.

CHEN Y L. Research on histogram equalization processing of gray image [D]. Hefei: Anhui University, 2014. (in Chinese)

[41] 寧媛, 李皖. 圖像去噪的幾種方法分析比較. 貴州工業(yè)大學(xué)學(xué)報(自然科學(xué)版), 2005, 34(4): 63-66.

NING Y, LI W. Analysis and comparison of some techniques on image denoising. Journal of Guizhou University of Technology (Natural Science Edition), 2005, 34(4): 63-66. (in Chinese)

[42] 楊光義. 圖像質(zhì)量評價及其在圖像去噪中的應(yīng)用研究[D]. 武漢: 武漢大學(xué), 2018.

YANG G Y. Image quality evaluation and its application in image denoising [D]. Wuhan: Wuhan University, 2018. (in Chinese)

[43] 苑麗紅, 付麗, 楊勇, 苗靜. 灰度共生矩陣提取紋理特征的實驗結(jié)果分析. 計算機應(yīng)用, 2009, 29(4): 1018-1021.

YUAN L H, FU L, YANG Y, MIAO J. Analysis of texture feature extracted by gray level co-occurrence matrix. Journal of Computer Applications, 2009, 29(4): 1018-1021. (in Chinese)

[44] 王璨, 李志偉. 利用融合高度與單目圖像特征的支持向量機模型識別雜草. 農(nóng)業(yè)工程學(xué)報, 2016, 32(15): 165-174.

WANG C, LI Z W. Weed recognition using SVM model with fusion height and monocular image features. Transactions of the Chinese Society of Agricultural Engineering, 2016, 32(15): 165-174. (in Chinese)

Potato Tuber Skin Roughness Classification Analysis Based on Image Characteristics Recognition

TANG ZhenSan1, YUAN JianLong1, KANG LiangHe2, CHENG LiXiang1, Lü Tai3, YANG Chen3, ZHANG Feng1

1College of Agriculture, Gansu Agricultural University/State Key Laboratory of Aridland Crop Science/Gansu Key Laboratory of Crop Improvement & Germplasm Enhancement, Lanzhou 730070;2College of Information Science and Technology, Gansu Agricultural University, Lanzhou 730070;3Tianshui Institute of Agricultural Sciences, Tianshui 741000, Gansu

【Objective】The classification analysis of potato tuber skin roughness could provide the non-destructive testing methods for tuber appearance quality traits, which would establish the theoretical and practical base for the objective evaluation of tuber quality and high-throughput screening varieties.【Method】Seventy-nine potato varieties (lines) were selected as materials, and the images of tuber skin with and without bud-eyes were taken by camera. The tuber skin images were preprocessed using MATLAB R2016a software. Eight materials were randomly selected to compare the effect of image graying, enhancement and denoising using the correlation function indicators. The image characteristic parameters, angular second moment (ASM), entropy (ENT), contrast (CON) and correlation (COR) were extracted using the gray level co-occurrence matrix (GLCM), and the suitable distance (d) of GLCM were determined. The differences in two types of tuber skin image feature parameters were compared, and the set of tuber skin image features with less difference was selected for statistical analysis and classification recognition. The support vector machine (SVM) and backpropagation neural network (BPNN) models were constructed for tuber skin roughness classification, and the evaluation indexes of model grading accuracy were accuracy, precision, recall and harmonic mean, respectively. 【Result】The texture structure of tuber skin image after grayscale processing using the weighted average method was clear, and the evaluation value of image clarity was 2.5698±0.5959, which was significantly higher than that of the mean method (1.8035±0.4856) and the maximum method (1.0535±0.4088). The gray scale range of tuber skin image after histogram equalization enhancement was expanded from 100-200 to 0-200, which made the gray distribution wider. The salt noise denoising effect of tuber skin images using the median filter under 3×3 sliding windows was obvious, and the peak signal-to-noise ratio (PSNR) was maximum (28.6250±3.9784 Bp), which was significantly higher than that under 3×3 and 5×5 windows. Two types of tuber skin image feature parameters extracted by GLCM (d=4) were significantly different, and the set of tuber skin image features (without bud-eyes) with less difference was selected for statistical analysis and classification recognition. The results indicated that the variation coefficient of these parameters was varied significantly. The variation coefficient of contrast was the largest (0.40), followed by the angular second moment (0.24) and correlation (0.23), and the variation coefficient of entropy was the smallest (0.18). Using the feature set as the input variable of tuber skin classification model, the overall classification performance of SVM was higher than BP neural network, and the accuracy reached 87.5%. Especially, the prediction accuracy and recognizability of SVM for smooth and heavy hemp skins was the highest. The accuracy reached 100%, the recall reached 85.7% and 100%, and the harmonic mean reached 100% and 92.3%, respectively. 【Conclusion】The combination of the image processing techniques presented in this study and the GLCM extracted texture feature parameters could effectively characterize potato tuber skin roughness variations. The tuber skin roughness grading based on machine vision could be achieved by constructing SVM classification model, and the accuracy reached 87.5%.

; tuber skin roughness; image characteristic; machine vision; support vector machine

10.3864/j.issn.0578-1752.2023.22.006

2023-03-17;

2023-07-24

國家重點研發(fā)計劃(SQ2022YFD1600328)、甘肅省科技重大專項(21ZD11NA002,21ZD11NA009)

唐振三,E-mail:1316740746@qq.com。通信作者張峰,E-mail:zhangf@gsau.edu.cn

(責(zé)任編輯 趙伶俐)

猜你喜歡
特征參數(shù)粗糙度紋理
故障診斷中信號特征參數(shù)擇取方法
基于特征參數(shù)化的木工CAD/CAM系統(tǒng)
基于BM3D的復(fù)雜紋理區(qū)域圖像去噪
軟件(2020年3期)2020-04-20 01:45:18
基于無人機影像的巖體結(jié)構(gòu)面粗糙度獲取
甘肅科技(2020年20期)2020-04-13 00:30:18
冷沖模磨削表面粗糙度的加工試驗與應(yīng)用
模具制造(2019年4期)2019-06-24 03:36:48
使用紋理疊加添加藝術(shù)畫特效
基于PSO-VMD的齒輪特征參數(shù)提取方法研究
TEXTURE ON TEXTURE質(zhì)地上的紋理
Coco薇(2017年8期)2017-08-03 15:23:38
基于BP神經(jīng)網(wǎng)絡(luò)的面齒輪齒面粗糙度研究
鋼材銹蝕率與表面三維粗糙度參數(shù)的關(guān)系
盐山县| 商河县| 玛曲县| 两当县| 武隆县| 吉水县| 岑巩县| 普兰店市| 香格里拉县| 合肥市| 郑州市| 库车县| 大安市| 永嘉县| 光泽县| 孟津县| 花垣县| 雅江县| 潮安县| 荆门市| 绥芬河市| 射阳县| 洛浦县| 博白县| 安庆市| 同心县| 罗定市| 南通市| 台北市| 开平市| 甘孜| 仪征市| 福鼎市| 大竹县| 沿河| 阿拉善盟| 咸丰县| 华宁县| 新营市| 永寿县| 安乡县|