基于激光诱导击穿光谱对长白山黑猪肉部位分类

猪肉内部结构成分复杂，各部位的成分相似，分辨较为困难。结合激光诱导击穿光谱技术，通过光谱分析的方式提高分类精度。以5种不同部位的长白山黑猪肉（里脊肉、梅花肉、后腿肉、前腿肉、五花肉）作为待测样品，通过冷藏、切片等预处理方法，探究激光诱导击穿光谱技术鉴别猪肉脂肪与肌肉及其不同部位的可行性。首先通过采集猪肉脂肪样品与肌肉样品的LIBS谱线信息发现，猪肉中Mg，K，Fe，Cu，Ca和Na等元素较为丰富，并在脂肪样品光谱中发现C-N键，与肌肉样品LIBS谱线信息相比，脂肪样品受其内部水分、有机质成分影响致使其谱线信息背景和噪声信号干扰较大，二者谱线信息存在一定差异，说明LIBS可对脂肪组织与肌肉组织进行鉴别。通过对目标元素Ca，Na，Mg，K和Al其LIBS特征谱线强度进行检测，计算Mg/Ca，Al/Ca，Na/Ca和K/Ca比值，发现与Al/Ca和Mg/Ca相比，Na/Ca和K/Ca各部位元素比值分布差异明显，在此基础上，根据Na/Ca和K/Ca比值，计算猪肉各部位元素分布决策阈值（1-α）=90%]。发现与Al/Ca和Mg/Ca比值相比，Na/Ca和K/Ca更能明显的反应出各部位元素分布的不同。其比值分布阈值基本可对猪肉各部位进行区分。以前腿肉与后腿肉为例，前腿肉Na/Ca和K/Ca比值分别分布在1.29～1.58和0.31～0.42，后腿肉Na/Ca和K/Ca比值分别分布在0.98～1.18和0.15～0.23。其元素比值分布无明显重叠。最后，为提高LIBS技术对猪肉不同组织分类的可靠性，将光谱元素强度比值数据与主成分分析法相结合，基本可以实现对猪肉各部位的分类，说明元素特征谱线强度比值在对猪肉各部位分类时具有一定的预测精度。该工作证明，使用激光诱导击穿光谱技术对猪肉进行分类识别等定性分析时具有一定可行性，有望适用于其他生物组织检测分析。

Classification of Changbai Mountains Pork Based on Laser-Induced Breakdown Spectroscopy

The internal structure of pork is complex,the matrix effect is strong,the components of each part are similar,so it is not easy to distinguish.Combined with laser-induced breakdown spectroscopy,the classification accuracy is improved by spectral analysis.Five different parts of Changbai Mountains Pork(tenderloin,plum blossom,hind leg,front leg and streaky pork)were used as the samples to be tested.The feasibility of identifying pork fat,muscle and their different parts by laser-induced breakdown spectroscopy was explored by means of cold storage,slice and other pretreatment methods.Firstly,by collecting LIBS spectral line information of fat pork samples and muscle samples,it is found that Mg,K,Fe,Cu,CA,Na and other elements are abundant in pork,and C—N bond is found in the spectrum of fat samples.Compared with LIBS spectral line information of muscle samples,the background and noise signal of the spectral line information of fat samples are greatly disturbed due to the influence of their internal moisture and organic matter composition.There are some differences,which indicates that LIBS can be used to distinguish adipose tissue from muscle tissue.Through the detection of the LIBS characteristic spectral intensity of the target elements Ca,Na,Mg,K and Al,the ratios of mg/Ca,Al/Ca,Na/Ca and K/Ca were calculated.It was found that the distribution of element ratio of Na/Ca and K/Ca was significantly different from that of Al/Ca and Mg/ca.On this basis,according to the ratio of Na/Ca and K/Ca,the decision threshold of element distribution of pork was calculated(1-α)=90%].It is found that Na/Ca and K/CA can reflect the distribution of elements more clearly than Al/Ca and Mg/ca.The threshold value of the ratio distribution can be used to distinguish different parts of pork.Taking the front leg meat and the back leg meat as an example,the Na/Ca and K/Ca ratios of the front leg meat were 1.29~1.58 and 0.31~0.42 respectively,and the Na/Ca and K/Ca ratios of the back leg meat were 0.98~1.18 and 0.15~0.23 respectively.There was no obvious overlap in the distribution of element ratio.Finally,in order to improve the reliability of LIBS technology in the classification of different tissues of pork,the spectral element intensity ratio data and principal component analysis method are combined,which can basically achieve the classification of various parts of pork,indicating that the element characteristic spectral line intensity ratio has certain prediction accuracy in the classification of various parts of pork.The whole work has proved that it is feasible to use laser-induced breakdown spectroscopy in qualitative analysis of pork,such as classification and identification,which is expected to be suitable for other biological tissue detection and analysis.