基于太赫兹反射成像技术的花生空果无损检测

目的 基于太赫兹反射成像技术(Terahertz Reflection Imaging Method, THz-RIM)实现花生空果无损检测，为太赫兹技术在农产品及食品快速无损检测中的应用奠定基础。方法 利用太赫兹脉冲光谱仪TeraPulse 4000及反射成像附件和采集软件，在0.2 mm采集步长下对去壳花生仁进行太赫兹反射成像预实验，确定出正式实验的图像颜色模式为hot模式。人为制备若干个花育36号满仁花生果样本和空果样本(仅有一粒花生仁)，在与预试验相同的条件下采集样本光谱图像，以时域/频域2点外相加成像(Integrated Time Slice)重构方式进行图像重构, 突出花生空果样本图像的显著几何特征。结果 通过太赫兹反射图像可以实现满仁与花生空果样本的检测区分，但图像细节仍然比较模糊，需进一步提高图像清晰度。采用数字图像方法进行图像增强，利用OpenCV-Python提高图像的亮度、色度、对比度及锐度。提升对比度后的图像可观测度最高，且空果图像中黑色部分面积占比最大，与含仁部分对比最为明显，可以实现空果及满仁花生果的快速区分。结论 探索结果表明，THz波对花生外壳具有良好的穿透性，利用THz-RIM技术结合OpenCV-Python数字图像处理方法进行花生空果无损检测具有可行性，利用太赫兹光谱技术对油料作物内部品质进行无损检测研究具有极大潜力。

Nondestructive inspection of empty peanut based on Terahertz reflection imaging method

Objective To inspect empty peanut nondestructively based on Terahertz reflection imaging method (THz-RIM). Methods TeraPulse 4000 Terahertz pulse spectrometer, reflection imaging accessories and the acquisition software of the instrument were used to conduct the pre-experiment of Terahertz reflection imaging on the shelled peanut kernels at the acquisition step of 0.2 mm, and the color mode of the image for the formal experiment was also identified to be hot mode. Several samples of Huayu No.36 which were full of kernel peanut fruit and empty fruit samples (with only one peanut kernel) were prepared artificially. Then the spectral images of the samples were collected under the same conditions as the pre-experiment. And the integrated time slice reconstruction method was used to reconstruct the image to highlight the significant geometric features of the empty peanut image. Results The Terahertz reflection image could be used to distinguish between full peanut and empty peanut, but the details of the image were still vague. There needed other ways to improve the clarity of the images that were collected in above experiment. For instance, in this paper, the digital image processing method was used for image enhancement. OpenCV-Python was selected to improve the brightness, chroma, contrast and sharpness of the image. The image with improved contrast had the highest observability, and the black part of the kernel part accounted for the largest proportion, which was most obvious compared with the empty fruit. That was to say full peanut and empty peanut could be distinguished in this way. Conclusion THz waves can penetrate peanut shell, and it is feasible to use THz-RIM technology combined with OpenCV-Python for nondestructive inspection of shelled empty peanut. Also, using Terahertz spectroscopy technology to conduct nondestructive inspection research on the quality of oil crop itself has great potential.