How Does Spectral Technology Make Fruit Quality Sorting Smarter?

水(shuǐ)果的(dí)品質分選直接決定了(liǎo)供(gōng)應(yīng)鏈的經(jīng)濟價(jià)值，精準識別(bié)內部損(sǔn)傷(shāng)不僅能顯著降(jiàng)低優質(zhì)果(guǒ)的損耗率，更(gēng)能(néng)提供穩定(dìng)可靠的高(gāo)品(pǐn)質(zhì)水果。然而(ér)，傳(chuán)統(tǒng)的人工分(fēn)選(xuǎn)和外部(bù)檢(jiǎn)測(cè)難以(yǐ)發現(xiàn)瘀(yū)傷，凍傷(shāng)，水(shuǐ)心病(bìng)等內(nèi)部缺陷(xiàn)，導致大(dà)量外(wài)表完(wán)好的(dí)水果因隱性損傷而被誤判。

光譜(pǔ)與高光譜成像(xiàng)技術的(dí)出(chū)現(xiàn)，讓水(shuǐ)果內(nèi)部品(pǐn)質(zhì)的(dí)無損(sǔn)檢(jiǎn)測成為(wéi)可能。這些(xiē)技(jì)術通(tōng)過解(jiě)析(xī)水(shuǐ)果內部水分，糖分及(jí)細(xì)胞結(jié)構(gòu)的特(tè)征光(guāng)信(xìn)號(hào)，實(shí)現對瘀(yū)傷(shāng)，凍傷(shāng)，海綿組織(zhī)病變(biàn)等隱(yǐn)性缺陷的“透視(shì)"，從而(ér)大幅提(tí)升分選(xuǎn)精(jīng)度(dù)。

Fruit quality sorting directly determines the economic value of the supply chain. Accurately identifying internal damage not only significantly reduces the loss rate of premium fruits but also provides stable and reliable high-quality fruits for the premium market. However, traditional manual sorting and external inspection struggle to detect internal defects such as bruises, frost damage, and watercore, leading to misjudgment of large quantities of outwardly intact fruits due to hidden damage.

The emergence of spectral and hyperspectral imaging technologies has made non-destructive testing of internal fruit quality possible. These technologies analyze characteristic optical signals related to internal moisture, sugar content, and cellular structure, enabling "visualization" of hidden defects like bruises, frost damage, and spongy tissue disorders, thereby significantly improving sorting accuracy.

水(shuǐ)果(guǒ)分選綫(xiàn)，圖(tú)片源(yuán)自網絡 / Fruit sorting line, image source: Internet

在(zài)國內外，科研團(tuán)隊(duì)已通過(guò)大(dà)量(liáng)實驗驗證了(liǎo)光譜技術在水果(guǒ)分選(xuǎn)中的(dí)潛力(lì)。例如，江(jiāng)蘇大學(xué)團隊(duì)利用高(gāo)光譜成(chéng)像係(xì)統檢測(cè)蘋果(guǒ)的輕微損傷，發現547nm波段的特征光譜能(néng)清晰反映皮下細胞損傷(shāng)，通(tōng)過主(zhǔ)成(chéng)分分析（PCA）提(tí)取該波段圖(tú)像，結(jié)合二(èr)次差(chà)分(fēn)算法(fǎ)消(xiāo)除果麵亮度不均(jūn)幹擾，最(zuì)終實現(xiàn)88.57%的(dí)損(sǔn)傷(shāng)識別(bié)率(shuài)。

Globally, research teams have validated the potential of spectral technology in fruit sorting through extensive experiments. For example, a team from Jiangsu University used a hyperspectral imaging system to detect slight damage in apples. They found that the characteristic spectrum at the 547nm wavelength clearly reflects subcutaneous cellular damage. By extracting images of this wavelength through principal component analysis (PCA) and combining it with a second-order difference algorithm to eliminate interference from uneven surface brightness, they achieved an 88.57% damage recognition rate.

蘋(píng)果(guǒ)的輕微(wēi)損傷(shāng)和(hé)正(zhèng)常(cháng)區域(yù)的光譜曲綫(xiàn)

Refectance spectra from the subtle bruise and normal region on the apple

類(lèi)似(sì)地(dì)，國(guó)外研(yán)究(jiū)團隊在芒果海綿組(zǔ)織檢(jiǎn)測(cè)中(zhōng)，通過(guò)優(yōu)化特定(dìng)波段(duàn)的(dí)Fisher特征選(xuǎn)擇算法(fǎ)，使(shǐ)分(fēn)類(lèi)準(zhǔn)確率達到84.5%，且預測缺(quē)陷位(wèi)置與實(shí)際(jì)損傷(shāng)的(dí)誤(wù)差小於(yú)1mm。

Similarly, a foreign research team optimized the Fisher feature selection algorithm for specific wavelengths in mango spongy tissue detection, achieving a classification accuracy of 84.5% with prediction errors of defect locations within 1mm of actual damage.

缺(quē)陷樣本與(yǔ)健康樣本(běn)的光譜圖，(a) 波(bō)長範(fàn)圍673nm–1100nm，(b) 波長範圍(wéi)1100nm–1900nm

A plot of defective and healthy samples  (a) Wavelength range 673 nm–1100 nm. (b) Wavelength range 1100 nm–1900 nm.

這些研(yán)究成果(guǒ)為實際產(chǎn)綫應(yīng)用(yòng)奠(diàn)定了基礎(chǔ)。基於光(guāng)譜的(dí)水(shuǐ)果分析(xī)係(xì)統通常由(yóu)光源模塊(kuài)，光譜(pǔ)儀(yí)/高光(guāng)譜成像(xiàng)儀，傳送帶(dài)等核(hé)心硬(yìng)件組成，其(qí)工(gōng)作(zuò)流程(chéng)包(bāo)括(kuò)樣(yàng)品采(cǎi)集(jí)，光(guāng)譜(pǔ)數據預(yù)處(chǔ)理，化(huà)學分(fēn)析方法(fǎ)測定(dìng)水(shuǐ)果(guǒ)樣(yàng)品成分的(dí)準確含(hán)量(liáng)，模型構建與驗證，優化(huà)模型等關(guān)鍵(jiàn)步(bù)驟(zhòu)。而在(zài)實(shí)際分選(xuǎn)場景中(zhōng)，這個流(liú)程如(rú)何高效(xiào)運(yùn)行？關(guān)鍵在(zài)於(yú)自動化和(hé)光譜(pǔ)檢(jiǎn)測的緊(jǐn)密結(jié)合：

1. 動態(tài)觸(chù)發：傳送帶水果(guǒ)抵達(dá)檢測位(wèi)，光(guāng)電傳(chuán)感器(qì)觸(chù)發(fā)光(guāng)源

2. 光譜采集：光(guāng)源(yuán)發(fā)射(shè)出光(guāng)，光譜儀獲反射(shè)/透射光(guāng)譜

3. 數據(jù)處理：光(guāng)譜(pǔ)儀分(fēn)解(jiě)特征(zhēng)峰，分析模型(xíng)實(shí)時(shí)輸出(chū)糖度(dù)/損傷(shāng)值

4. 分(fēn)揀(jiǎn)執(zhí)行：觸發品質分級(jí)

These research findings lay the foundation for practical production line applications. Spectral-based fruit analysis systems typically consist of core hardware such as a light source module, spectrometer/hyperspectral imager, and conveyor belt. The workflow includes key steps such as sample collection, spectral data preprocessing, chemical analysis to determine the accurate content of fruit sample components, model construction and validation, and model optimization. In actual sorting scenarios, the efficiency of this process hinges on the seamless integration of automation and spectral detection:

1. Dynamic Triggering: Photoelectric sensors activate the light source when fruit reaches the detection position on the conveyor belt.

2. Spectral Acquisition: The light source emits light, and the spectrometer captures the reflected/transmitted spectra.

3. Data Processing: The spectrometer decomposes characteristic peaks, and the analysis model outputs real-time brix/damage values.

4. Sorting Execution: Quality grading is triggered for sorting.

高光譜(pǔ)係統(tǒng)的示(shì)意圖（由於水果的尺(chǐ)寸(cùn)大小，果肉薄厚，糖(táng)酸(suān)度高有低，且(qiě)分布不均(jūn)的(dí)情況，光譜(pǔ)采集時光源擺放有(yǒu)多(duō)種方(fāng)式(shì)）

Schematic diagram of a hyperspectral system (Due to variations in fruit size, flesh thickness, and uneven distribution of sugar/acid content, multiple light source configurations are used during spectral acquisition)

目前，光纖(xiān)光譜(pǔ)儀因其(qí)成本低，結構緊(jǐn)湊等(děng)優勢(shì)，仍是水(shuǐ)果(guǒ)分選(xuǎn)的(dí)主(zhǔ)流(liú)設備。但對於(yú)聖(shèng)女果，櫻(yīng)桃等小(xiǎo)尺(chǐ)寸(cùn)水(shuǐ)果(guǒ)，光(guāng)纖光譜(pǔ)儀的(dí)檢測效(xiào)率可能受(shòu)限，而高光譜成像(xiàng)儀(yí)憑借其(qí)空(kōng)間(jiān)與(yǔ)光(guāng)譜信(xìn)息(xī)的同步獲(huò)取(qǔ)能力(lì)，理論上(shàng)能(néng)實現(xiàn)更(gēng)高效(xiào)的分(fēn)選。然而(ér)，高光(guāng)譜(pǔ)設備(bèi)的(dí)成本和數據處(chǔ)理(lǐ)復雜(zá)度(dù)仍(réng)是實(shí)際應用(yòng)中的挑戰。

未來(lái)，隨著(zhuó)硬(yìng)件(jiàn)優化和算法(fǎ)的持(chí)續升級(jí)，光譜(pǔ)技術有(yǒu)望(wàng)在更(gēng)多水(shuǐ)果(guǒ)品類(lèi)中(zhōng)實現(xiàn)高(gāo)效(xiào)，經濟的分(fēn)選方(fāng)案，推(tuī)動水果供(gōng)應鏈向更(gēng)智(zhì)能(néng)，更精(jīng)準的(dí)方向發展。

Currently, fiber optic spectrometers remain the mainstream equipment for fruit sorting due to their low cost and compact structure. However, for small-sized fruits like cherry tomatoes and cherries, the detection efficiency of fiber optic spectrometers may be limited. In contrast, hyperspectral imagers, with their ability to simultaneously capture spatial and spectral information, theoretically enable more efficient sorting. Nevertheless, the cost of hyperspectral equipment and the complexity of data processing remain challenges in practical applications.

Looking ahead, with continuous hardware optimization and algorithm advancements, spectral technology is expected to deliver efficient and cost-effective sorting solutions for more fruit varieties, driving the fruit supply chain toward smarter and more precise development.

案(àn)例來源 / Source：

1. Zhao, J.-W., Liu, J.-H., Chen, Q.-S., & Vittayapadung, S. (2008). 利用高(gāo)光譜圖像(xiàng)技(jì)術(shù)檢測(cè)水果(guǒ)輕微損傷(shāng) [Detection of slight fruit bruises using hyperspectral imaging technology]. Transactions of the Chinese Society for Agricultural Machinery, 39(1), 106-109.

2. Raghavendra, A., Guru, D. S., & Rao, M. K. (2021). Mango internal defect detection based on optimal wavelength selection method using NIR spectroscopy. Artificial Intelligence in Agriculture, 5, 43-51.