當我們描(miáo)述(shù)一(yī)杯(bēi)咖啡(fēi)帶(dài)有花(huā)香，堅果(guǒ)或(huò)焦(jiāo)糖風(fēng)味(wèi)時，是否想過，這(zhè)份(fèn)獨特的風味能否(fǒu)在不依賴杯(bēi)測的(dí)情況(kuàng)下(xià)，於(yú)烘(hōng)焙前就(jiù)能被預測(cè)？

高光譜(pǔ)成像技(jì)術正(zhèng)在(zài)將(jiāng)這種構想變(biàn)為可能。高光譜(pǔ)能(néng)無損地(dì)掃(sǎo)描(miáo)咖啡生豆，獲得其完整的光(guāng)譜(pǔ)數據，研發人員從(cóng)光譜中找出與(yǔ)特(tè)定風(fēng)味物質（如糖分，caffeine）之間(jiān)的(dí)定量關(guān)係，以此為基礎(chǔ)，構建出可靠的預測(cè)模型(xíng)。通過(guò)這些模(mó)型(xíng)，我們得(dé)以科學(xué)地預(yù)見(jiàn)咖啡豆(dòu)的(dí)風味輪(lún)廓(kuò)。

在(zài)一項研(yán)究中，科研人員將單顆(kē)咖(kā)啡豆(dòu)（生(shēng)豆(dòu)）放(fàng)置在(zài)黑色樣(yàng)品(pǐn)台上(shàng)，利用高(gāo)光譜相機對(duì)其進行(háng)掃描。高(gāo)光譜(pǔ)相機的(dí)覆蓋短波紅(hóng)外(wài)區域（900-2500nm），咖啡(fēi)豆內(nèi)部(bù)的主(zhǔ)要化(huà)學(xué)成分(fēn)，如糖(táng)分(fēn)，生(shēng)物(wù)鹼(jiǎn)，油脂等(děng)，在此波(bō)段(duàn)下會呈現出(chū)獨(dú)特的“光譜指紋(wén)"。利(lì)用專(zhuān)業(yè)的軟(ruǎn)件分析(xī)和建(jiàn)立(lì)模型，定量(liáng)推測(cè)豆(dòu)子中各種(zhǒng)物質(zhì)的含量。

研究(jiū)人員(yuán)掃描了(liǎo)多個(gè)產區的數百顆樣(yàng)本，以檢測(cè)蔗糖（甜味），caffeine（苦(kǔ)味(wèi)）和(hé)葫蘆巴鹼(jiǎn)（烘焙(bèi)香氣）這三(sān)種(zhǒng)關鍵風味前(qián)體(tǐ)。

通(tōng)過結合高(gāo)光譜(pǔ)數據與(yǔ)液相(xiāng)色(sè)譜質譜法(fǎ)的精密(mì)測量(liáng)值(zhí)，應用(yòng)PLSR算法(fǎ)，研究團(tuán)隊(duì)構(gòu)建了上述成(chéng)分的定(dìng)量(liáng)預測(cè)模型(xíng)。交叉(chā)驗(yàn)證結(jié)果(guǒ)表明(míng)，模型對caffeine和葫蘆巴鹼的(dí)預測精度很高（R² > 0.8），對(duì)蔗糖的(dí)預測(cè)也可用於初(chū)步(bù)篩(shāi)選(xuǎn)。

利用(yòng)高光譜成像(xiàng)的特(tè)性(xìng)，團(tuán)隊還生(shēng)成了(liǎo)這(zhè)些(xiē)化(huà)學成分在豆子內部的空(kōng)間分(fēn)布(bù)圖，直觀顯(xiǎn)示了它(tā)們(mén)的(dí)不(bù)均(jūn)匀(yún)分布，這(zhè)為了解咖啡(fēi)豆的生(shēng)理(lǐ)結(jié)構提(tí)供了(liǎo)新(xīn)視角。

通過高光(guāng)譜成(chéng)像（HSI）獲取的純品(pǐn)參(cān)比(bǐ)物(wù)質（caffeine，蔗糖(táng)和(hé)葫蘆巴鹼(jiǎn)）以(yǐ)及(jí)磨碎生咖啡(fēi)豆樣(yàng)品的平均光譜，同(tóng)時(shí)展示了(liǎo)1400nm單一(yī)光(guāng)譜波(bō)段的吸光(guāng)度(dù)圖(tú)像(xiàng)（右(yòu)側(cè)）

該研究團隊還(huán)轉向烘焙後(hòu)咖啡(fēi)豆(dòu)，為(wéi)應對更復雜的風味分析(xī)，采用(yòng)了(liǎo)可(kě)同時預測(cè)多個響應(yīng)變量(liáng)的(dí)PLS2算(suàn)法。研究人(rén)員不僅建立(lì)了針(zhēn)對醛(quán)類，吡(pǐ)嗪(qín)類(lèi)等(děng)化(huà)學族群(qún)的預(yù)測模(mó)型(xíng)，更進一步(bù)將化合物(wù)按(àn)其感(gǎn)官(guān)特征(zhēng)（如堅(jiān)果(guǒ)香(xiāng)，甜香）分組，建(jiàn)立(lì)了(liǎo)針對整體(tǐ)風味屬性的模型。結(jié)果表(biǎo)明，對(duì)醛類（甜(tián)香(xiāng)）和(hé)吡嗪(qín)類（烘烤(kǎo)香(xiāng)）等(děng)族群(qún)的(dí)預測效果尤(yóu)為出色(sè)。

為(wéi)了驗(yàn)證這一預(yù)測能力的(dí)實(shí)際(jì)價(jià)值(zhí)，研究人(rén)員進行(háng)了一項(xiàng)實驗(yàn)：他們利(lì)用建立(lì)好的(dí)模(mó)型(xíng)，對(duì)一(yī)批(pī)阿拉伯比卡(qiǎ)咖啡(fēi)豆進行掃(sǎo)描，並(bìng)根據模型(xíng)預(yù)測的吡(pǐ)嗪(qín)含(hán)量和堅(jiān)果香氣強度(dù)，手(shǒu)動篩選出(chū)預(yù)測值(zhí)max和min的10%的(dí)豆子，分(fēn)別組成新的(dí)批次。

對這(zhè)兩(liǎng)個(gè)批次豆(dòu)子的化學分析證(zhèng)實，分選效(xiào)果極其顯著(zhù)。被預測為(wéi)“高吡(pǐ)嗪"的(dí)批次，其實(shí)際吡(pǐ)嗪(qín)類物(wù)質含量(liáng)顯著高於原(yuán)始混合批次(cì)和“低吡(pǐ)嗪"批次(cì)。

這(zhè)證(zhèng)明，高光譜成像技術結合(hé)預測模型，能夠(gòu)有效(xiào)識別咖啡(fēi)豆內(nèi)部(bù)的特定風味物質(zhì)差異。該研(yán)究(jiū)為(wéi)在(zài)產業(yè)綫上(shàng)實現(xiàn)非(fēi)破壞(huài)性的(dí)精(jīng)準風味(wèi)分(fēn)選(xuǎn)提供(gōng)了新思路，展現(xiàn)了(liǎo)其用於(yú)生產(chǎn)風味(wèi)定製(zhì)化(huà)咖(kā)啡產品的技術(shù)潛力(lì)，但其(qí)大規(guī)模穩(wěn)定應(yīng)用的效能仍有待(dài)進(jìn)一(yī)步(bù)驗(yàn)證(zhèng)。

按A.預測吡嗪(qín)類(lèi)化合物(wù)含量或(huò)B.分析(xī)預測 堅(jiān)果味（粗體標(biāo)注）的前麵(miàn)10%（高含(hán)量組，H）或後10%（低含量組(zǔ)，L）對(duì)咖(kā)啡(fēi)豆進行(háng)分選(xuǎn)後，分選(xuǎn)試驗(yàn)對(duì) 4 組(zǔ)揮發(fā)性化合物（吡(pǐ)嗪(qín)類(lèi)，醛(quán)類(lèi)，酮(tóng)類和雜(zá)環(huán)含氮(dàn)化(huà)合物(wù)）相對(duì)豐度(dù)及分析(xī)預測(cè)的堅(jiān)果味(wèi)，果香味，酸味和烘(hōng)焙(bèi)味的(dí)影響。

從咖(kā)啡生(shēng)豆(dòu)的檢測，到(dào)熟(shú)豆後天風(fēng)味(wèi)的預測(cè)，高光譜(pǔ)成像技術為我(wǒ)們(mén)提供了一(yī)條貫穿咖(kā)啡品質管控全程的(dí)*紐帶。它(tā)讓“看豆(dòu)識風(fēng)味"成為可能，將咖啡(fēi)的品質控製從傳統(tǒng)依(yī)賴(lài)經驗(yàn)的“批(pī)量(liáng)評(píng)估(gū)"，推(tuī)向了一(yī)個(gè)數字(zì)化，可視化的“單(dān)顆精(jīng)準管(guǎn)理(lǐ)"新時代。

案(àn)例(lì)來源(yuán)：

Caporaso, N., Whitworth, M. B., Grebby, S., & Fisk, I. D. (2018). Non-destructive analysis of sucrose, caffeine and trigonelline on single green coffee beans by hyperspectral imaging. Food Research International, 106, 193–203.

Caporaso, N., Whitworth, M. B., & Fisk, I. D. (2022). Prediction of coffee aroma from single roasted coffee beans by hyperspectral imaging. Food Chemistry, 371, 131159.

Can Hyperspectral Imaging Detect the Flavor of Coffee Beans?

When describing a cup of coffee as having floral, nutty, or caramel notes, have you ever wondered whether these unique flavors could be predicted before roasting—without relying on cupping?

Hyperspectral imaging technology is turning this concept into a reality. By scanning green coffee beans non-destructively, hyperspectral systems capture complete spectral data. Researchers identify quantitative relationships between the spectral signatures and specific flavor compounds—such as sugars and caffeine—and use these to build reliable predictive models. Through these models, the flavor profile of coffee beans can be scientifically anticipated.

In one study, researchers placed individual green coffee beans on a black sample stage and scanned them using a hyperspectral camera. The camera covered the short-wave infrared range (900–2500 nm), where key chemical components inside the beans—such as sugars, alkaloids, and oils—exhibit distinct "spectral fingerprints." Using specialized software for analysis and modeling, the content of various compounds in the beans was quantitatively estimated.

Hundreds of samples from multiple global growing regions were scanned to detect three key flavor precursors: sucrose (sweetness), caffeine (bitterness), and trigonelline (roasty aroma).

By combining hyperspectral data with precise measurements from liquid chromatography–mass spectrometry, the research team applied PLSR (Partial Least Squares Regression) to develop quantitative prediction models for these components. Cross-validation results showed high prediction accuracy for caffeine and trigonelline (R² > 0.8), and sucrose predictions were suitable for preliminary screening.

Leveraging the capabilities of hyperspectral imaging, the team also generated spatial distribution maps of these chemical compounds inside the beans. These visualizations clearly revealed their uneven distribution, offering new insights into the physiological structure of coffee beans.

The average spectra of pure reference substances (caffeine, sucrose, and trigonelline) and ground green coffee bean samples obtained through hyperspectral imaging (HSI), along with the absorbance image at a single spectral band of 1400 nm (shown on the right).

The research team also examined roasted coffee beans. To address more complex flavor analysis, they employed the PLS2 algorithm, which can predict multiple response variables simultaneously. The researchers not only built prediction models for chemical groups such as aldehydes and pyrazines, but also grouped compounds by sensory attributes—such as nutty aroma and sweet aroma—to develop models targeting overall flavor characteristics. Results indicated particularly strong predictive performance for groups like aldehydes (sweet aroma) and pyrazines (roasty aroma).

To test the practical value of this predictive capability, the researchers conducted an experiment: using the established model, they scanned a batch of Arabica coffee beans. Based on the predicted pyrazine content and nutty aroma intensity, they manually selected the top 10% and bottom 10% of beans to form two new batches.

Chemical analysis of these two batches confirmed highly significant sorting results. The batch predicted as "high pyrazine" showed substantially higher actual pyrazine content compared to the original mixed batch and the "low pyrazine" batch.

This demonstrates that hyperspectral imaging combined with predictive models can effectively identify differences in specific flavor compounds within coffee beans. The study offers a new approach for non-destructive, precise flavor sorting on industrial production lines, highlighting the technology’s potential for producing customized coffee products. However, the efficiency and stability of large-scale application still require further validation.

After sorting coffee beans into either the top 10% (high-content group, H) or the bottom 10% (low-content group, L) based on A. predicted pyrazine content or B. predicted "nutty flavor" (indicated in bold), the sorting experiment’s impact on the relative abundance of four groups of volatile compounds (pyrazines, aldehydes, ketones, and nitrogen-containing heterocycles) and the predicted intensities of nutty flavor, fruity flavor, acidity, and roasty flavor.

From detecting green coffee beans to predicting the developed flavors of roasted beans, hyperspectral imaging provides a powerful link throughout the entire coffee quality control process. It makes it possible to "see flavor in the bean," shifting coffee quality control from traditional, experience-based "batch assessment" toward a new era of digital, visual, and "single-bean precision management."

Sources：

Caporaso, N., Whitworth, M. B., Grebby, S., & Fisk, I. D. (2018). Non-destructive analysis of sucrose, caffeine and trigonelline on single green coffee beans by hyperspectral imaging. Food Research International, 106, 193–203.

Caporaso, N., Whitworth, M. B., & Fisk, I. D. (2022). Prediction of coffee aroma from single roasted coffee beans by hyperspectral imaging. Food Chemistry, 371, 131159.