ALGORITHM DEVELOPMENT WITH VISIBLE/NEAR-INFRARED SPECTRA FOR DETECTION OF POULTRY FECES AND INGESTA

ALGORITHM DEVELOPMENT WITH VISIBLE/NEAR-INFRARED SPECTRA FOR DETECTION OF POULTRY FECES AND INGESTA
复制标题

用于检测家禽粪便和摄入量的可见光/近红外光谱的算法开发

DOI：
10.13031/2013.15629
复制
发表时间：
2003
期刊：
影响因子：
--
通讯作者：
P. Feldner
P. Feldner
中科院分区：
--
文献类型：
--
作者：
W. R. Windham;D. P. Smith;B. Park;K. Lawrence;P. Feldner

文献摘要

被引文献

相似文献

美国农业部农业研究局开发了一种方法和高光谱成像系统来检测 粪便（来自十二指肠、盲肠和结肠）和家禽尸体上的摄入物。该方法首先涉及使用多元 对粪便和未污染皮肤样本的可见/近红外 (Vis/NIR) 反射光谱进行数据分析 污染物的分类和关键波长的选择。四个主波长（434、517、565 和 628 nm） 通过主成分（PC）负载重量的强度来确定。关键波长在高光谱上进行了验证 受污染的肉鸡尸体的图像。具体来说，当商为 565 nm/517 nm 时，100% 的粪便污染物 在以玉米/豆粕饲料喂养的有限数量的肉鸡中检测到。这项研究的目的是验证 565 nm/517 nm 商用于对饲喂玉米、高粱或小麦饲料的肉鸡粪便/摄入物中未受污染的皮肤进行分类 并研究使用单项线性回归（STLR）来选择分类的关键波长。粪便（N = 369) 和未受污染的肉鸡胸皮肤 (N = 96) 在 440 至 880 nm 范围内进行分析。检测总体准确率 具有 565 nm/517 nm 商的任何类型饲料的污染率为 99%，其中 16 个未受污染的皮肤样本已分类 污染（误报）。 STLR 优化了 574 nm/588 nm 的新商，可对污染物进行 100% 分类 正确无误报。分母从 517 nm 到 588 nm 的变化可能是由于粪便颜色变大 与饲喂小麦或高粱的肉鸡的变异。此外，除以 588 nm 可以最大限度地减少亮度 (L*) 对 分类。使用 STLR 扫描光谱数据来查找与因变量相关的波长是 根据 PC 负载重量的强度选择关键波长的替代方案。尽管来自 Vis/NIR 的模型 光谱学和 STLR 表现良好，需要在未受污染和未受污染的高光谱图像上进行验证 受污染的尸体。
The USDA Agricultural Research Service has developed a method and a hyperspectral imaging system to detect feces (from duodenum, ceca, and colon) and ingesta on poultry carcasses. The method first involves the use of multivariate data analysis on visible/near-infrared (Vis/NIR) reflectance spectra of fecal and uncontaminated skin samples for classification of contaminates and selection of key wavelengths. Four dominant wavelengths (434, 517, 565, and 628 nm) were identified by intensity of principal component (PC) loading weights. Key wavelengths were validated on hyperspectral images of contaminated broiler carcasses. Specifically, with a quotient of 565 nm/517 nm, 100% of the fecal contaminates were detected in a limited population of broilers fed a corn/soybean meal diet. The objectives of this research was to validate the 565 nm/517 nm quotient to classify uncontaminated skin from feces/ingesta with broilers fed corn, milo, or wheat diets and to investigate the use of single-term linear regression (STLR) to select key wavelengths for classification. Feces (N = 369) and uncontaminated broiler breast skin (N = 96) were analyzed from 440 to 880 nm. The overall accuracy of detecting contamination for any type of feed with the 565 nm/517 nm quotient was 99% with 16 uncontaminated skin samples classified as contaminates (false positive). STLR optimized a new quotient of 574 nm/588 nm, which classified 100% of contaminates correctly with no false positives. The shift in the denominator from 517 to 588 nm is possibly due to greater fecal color variation from broilers fed wheat or milo. In addition, dividing by 588 nm minimized the effect of lightness (L*) on classification. The use of the STLR to scan the spectral data to find wavelengths correlated with the dependent variable is an alternative to selecting key wavelengths based on the intensity of PC loading weights. Although models from Vis/NIR spectroscopy and STLR performed well, they need to be validated on hyperspectral images of uncontaminated and contaminated carcasses.