In order to overcome the shortcomings that the reconstructed spectral reflectance may be negative when using the classic principal component analysis （PCA）to reduce the dimensions of the multi-spectral data, a nonne...In order to overcome the shortcomings that the reconstructed spectral reflectance may be negative when using the classic principal component analysis （PCA）to reduce the dimensions of the multi-spectral data, a nonnegative constrained principal component analysis method is proposed to construct a low-dimensional multi-spectral space and accomplish the conversion between the new constructed space and the multispectral space. First, the reason behind the negative data is analyzed and a nonnegative constraint is imposed on the classic PCA. Then a set of nonnegative linear independence weight vectors of principal components is obtained, by which a lowdimensional space is constructed. Finally, a nonlinear optimization technique is used to determine the projection vectors of the high-dimensional multi-spectral data in the constructed space. Experimental results show that the proposed method can keep the reconstructed spectral data in [ 0, 1 ]. The precision of the space created by the proposed method is equivalent to or even higher than that by the PCA.展开更多
The demarcation line between the cancerous lesion and the surrounding area could be easily recognized with flexible spectral imaging color enhancement (FICE) system compared with conventional white light images. The c...The demarcation line between the cancerous lesion and the surrounding area could be easily recognized with flexible spectral imaging color enhancement (FICE) system compared with conventional white light images. The characteristic f inding of depressed-type early gastric cancer (EGC) in most cases was revealed as reddish lesions distinct from the surrounding yellowish non-cancerous area without magnification. Conventional endoscopic images provide little information regarding depressed lesions located in the tangential line, but FICE produces higher color contrast of such cancers. Histological f indings in depressed area with reddish col- or changes show a high density of glandular structure and an apparently irregular microvessel in intervening parts between crypts, resulting in the higher color con- trast of FICE image between cancer and surrounding area. Some depressed cancers are shown as whitish lesion by conventional endoscopy. FICE also can pro- duce higher color contrast between whitish cancerous lesions and surrounding atrophic mucosa. For nearly flat cancer, FICE can produce an irregular structuralpattern of cancer distinct from that of the surrounding mucosa, leading to a clear demarcation. Most elevated-type EGCs are detected easily as yellowish lesions with clearly contrasting demarcation. In some cases, a partially reddish change is accompanied on the tumor surface similar to depressed type cancer. In addition, the FICE system is quite useful for the detection of minute gastric cancer, even without magnif ication. These new contrasting images with the FICE system may have the potential to increase the rate of detection of gastric cancers and screen for them more effectively as well as to determine the extent of EGC.展开更多
AIM:To conduct a preliminary study on the effect of flexible spectral imaging color enhancement (FICE) used in combination with ultraslim endoscopy by focusing on the enhanced contrast between tumor and non-tumor lesi...AIM:To conduct a preliminary study on the effect of flexible spectral imaging color enhancement (FICE) used in combination with ultraslim endoscopy by focusing on the enhanced contrast between tumor and non-tumor lesions. METHODS: We examined 50 lesions of 40 patients with epithelial tumors of the upper gastrointestinal tract before endoscopic submucosal dissection using ultraslim endoscopy with conventional natural color imag ing and with FICE imaging. We retrospectively invest igated the effect of the use of FICE on endoscopic diagn osis in comparison with normal light. RESULTS: Visibility of the epithelial tumors of the upper gastrointestinal tract with FICE was superior to normal light in 54% of the observations and comparable to normal light in 46% of the observations. There was no lesion for which visibility with FICE was inferior to that with normal light. FICE visualized 69.6% of hyperemic lesions and 58.8% of discolored lesions better than conventional endoscopy with natural color imaging. FICE sign if icantly improved the visibility of lesions with hyp ere mia or discoloration compared with normocolored lesions. CONCLUSION: This study suggests that the use of FICE would improve the ability of ultraslim endoscopy to detect epithelial tumors of the upper gastrointestinal tract.展开更多
The cross-spectral estimation methods are efficient in estimating the parameters of sinusoidal signals embedded in colored noise. But up to now, only FPT and cross-periodogram methods are used in this field, the moder...The cross-spectral estimation methods are efficient in estimating the parameters of sinusoidal signals embedded in colored noise. But up to now, only FPT and cross-periodogram methods are used in this field, the modern auto-spectral estimation method is introduced into cross-spectral estimation in this paper, meanwhile the cross-correlation based Yule-Walker equation is proposed theoretically and the moment and singular-value decomposition (SVD)) algorithms for cross-spectral estimation have been developed. Finally, a numerical example is given for comparing the presented methods with the well-known Cadzow’s SVD method.展开更多
【目的】叶面积指数(leaf area index,LAI)是表征作物长势、光合、蒸腾的重要指标。论文旨在研究不同生育期、多生育期无人机多光谱数据棉花LAI估测模型,明确不同生育期间棉花LAI估测模型变化规律,为实时掌握棉花长势并因地制宜进行田...【目的】叶面积指数(leaf area index,LAI)是表征作物长势、光合、蒸腾的重要指标。论文旨在研究不同生育期、多生育期无人机多光谱数据棉花LAI估测模型,明确不同生育期间棉花LAI估测模型变化规律,为实时掌握棉花长势并因地制宜进行田间科学管理提供依据。【方法】利用大疆精灵4多光谱无人机获取棉花现蕾期、初花期、结铃期、吐絮期多光谱图像和RGB图像。选用归一化差植被指数(NDVI)、绿度归一化差植被指数(GNDVI)、归一化差红边指数(NDRE)、叶片叶绿素指数(LCI)、优化的土壤调节植被指数(OSAVI)5种多光谱指数和修正红绿植被指数(MGRVI)、红绿植被指数(GRVI)、绿叶指数(GLA)、超红指数(EXR)、大气阻抗植被指数(VARI)5种颜色指数分别建立棉花各生育期及棉花生长多生育期数据集合,结合打孔法获取地面LAI实测数据,使用机器学习算法中偏最小二乘(PLSR)、岭回归(RR)、随机森林(RF)、支持向量机(SVM)、神经网络(BP)构建棉花LAI预测模型。【结果】覆膜棉花LAI随着生育期的变化呈现先增长后下降的趋势,现蕾期、初花期、结铃期内侧棉花叶面积指数均值均显著大于外侧(P<0.05);选择的指数在各时期彼此间均呈显著相关(P<0.05),总体而言,多光谱指数与颜色指数间的相关性随着生育期的进行而呈现下降趋势,选择的指数在各时期均与棉花LAI相关性显著(P<0.05),多光谱指数相关系数介于0.35—0.85,颜色指数相关系数介于0.49—0.71,相关系数绝对值较大的指数多为多光谱指数,颜色指数与棉花LAI的相关系数绝对值较小;估测模型性能结果显示棉花各生育期模型中多光谱指数优于颜色指数,且各指数模型预测性能随着生育期的变化呈现一定规律性,NDVI是预测棉花LAI的最优指数。从模型结果上看,RF模型和BP模型在各生育期下获得了较高的估计精度。初花期LAI反演模型精度最高,最优模型验证集R2为0.809,MAE为0.288,NRMSE为0.120。多生育期最优模型验证集R2为0.386,MAE为0.700,NRMSE为0.198。【结论】棉花内外侧LAI在现蕾期、初花期、结铃期存在显著差异。在各生育期中,RF和BP模型是预测棉花LAI较优模型。NDVI在各指数中表现最好,是预测棉花LAI的最优指数。多生育期模型效果较单生育期明显下降,最优指数为GNDVI,最优模型为BP。本研究中预测棉花LAI的最优窗口期是初花期。研究结果可为无人机遥感监测棉花LAI提供理论依据和技术支持。展开更多
目的利用近红外光谱及光谱融合策略,结合化学计量学方法,建立水性油墨的颜色预测模型,实现水性油墨印刷品颜色准确预测。方法采集不同酒精含量和不同调色墨含量的油墨的近红外光谱反射率和吸光度数据,并测得对应的印刷品的Lab值,然后建...目的利用近红外光谱及光谱融合策略,结合化学计量学方法,建立水性油墨的颜色预测模型,实现水性油墨印刷品颜色准确预测。方法采集不同酒精含量和不同调色墨含量的油墨的近红外光谱反射率和吸光度数据,并测得对应的印刷品的Lab值,然后建立单一光谱不同预处理过后的偏最小二乘(Partial Least Squares,PLS)模型,以及基于数据层融合和特征层融合的PLS模型,最终通过比较预测集决定系数和预测集均方根误差(Root Mean Square Error of Prediction,RMSEP)以及色差来评估模型的预测效果。结果单光谱建模,基于反射率建立的模型准确率高于基于吸光度建立的模型;数据层融合缺乏稳定性,对L和b值的预测有所提升,对a值的预测几乎不变;特征层融合建模效果明显好于单一光谱和数据层融合,对Lab的预测决定系数分别达到了0.9961、0.9939、0.9974;RMSEP值分别为0.1421、0.2126、0.2072;预测值与真实值的最大色差为0.6783。结论通过光谱特征融合技术能提高油墨颜色预测精度,准确预测出油墨颜色变化。展开更多
Recently,image-enhanced endoscopy(IEE) has been used to diagnose gastrointestinal tumors.This method is a change from conventional white-light(WL) endoscopy without dyeing solution,requiring only the push of a button....Recently,image-enhanced endoscopy(IEE) has been used to diagnose gastrointestinal tumors.This method is a change from conventional white-light(WL) endoscopy without dyeing solution,requiring only the push of a button.In IEE,there are many advantages in diagnosis of neoplastic tumors,evaluation of invasion depth for cancerous lesions,and detection of neoplastic lesions.In narrow band imaging(NBI) systems(Olympus Medical Co.,Tokyo,Japan),optical filters that allow narrow-band light to pass at wavelengths of 415 and 540 nm are used.Mucosal surface blood vessels are seen most clearly at 415 nm,which is the wavelength that corresponds to the hemoglobin absorption band,while vessels in the deep layer of the mucosa can be detected at 540 nm.Thus,NBI also can detect pit-like structures named surface pattern.The flexible spectral imaging color enhancement(FICE) system(Fujifilm Medical Co.,Tokyo,Japan) is also an IEE but different to NBI.FICE depends on the use of spectral-estimation technology to reconstruct images at different wavelengths based on WL images.FICE can enhance vascular and surface patterns.The autofluorescence imaging(AFI) video endoscope system(Olympus Medical Co.,Tokyo,Japan) is a new illumination method that uses the difference in intensity of autofluorescence between the normal area and neoplastic lesions.AFI light comprises a blue light for emitting and a green light for hemoglobin absorption.The aim of this review is to highlight the efficacy of IEE for diagnosis of colorectal tumors for endoscopic treatment.展开更多
基金The Pre-Research Foundation of National Ministries andCommissions (No9140A16050109DZ01)the Scientific Research Program of the Education Department of Shanxi Province (No09JK701)
文摘In order to overcome the shortcomings that the reconstructed spectral reflectance may be negative when using the classic principal component analysis （PCA）to reduce the dimensions of the multi-spectral data, a nonnegative constrained principal component analysis method is proposed to construct a low-dimensional multi-spectral space and accomplish the conversion between the new constructed space and the multispectral space. First, the reason behind the negative data is analyzed and a nonnegative constraint is imposed on the classic PCA. Then a set of nonnegative linear independence weight vectors of principal components is obtained, by which a lowdimensional space is constructed. Finally, a nonlinear optimization technique is used to determine the projection vectors of the high-dimensional multi-spectral data in the constructed space. Experimental results show that the proposed method can keep the reconstructed spectral data in [ 0, 1 ]. The precision of the space created by the proposed method is equivalent to or even higher than that by the PCA.
文摘The demarcation line between the cancerous lesion and the surrounding area could be easily recognized with flexible spectral imaging color enhancement (FICE) system compared with conventional white light images. The characteristic f inding of depressed-type early gastric cancer (EGC) in most cases was revealed as reddish lesions distinct from the surrounding yellowish non-cancerous area without magnification. Conventional endoscopic images provide little information regarding depressed lesions located in the tangential line, but FICE produces higher color contrast of such cancers. Histological f indings in depressed area with reddish col- or changes show a high density of glandular structure and an apparently irregular microvessel in intervening parts between crypts, resulting in the higher color con- trast of FICE image between cancer and surrounding area. Some depressed cancers are shown as whitish lesion by conventional endoscopy. FICE also can pro- duce higher color contrast between whitish cancerous lesions and surrounding atrophic mucosa. For nearly flat cancer, FICE can produce an irregular structuralpattern of cancer distinct from that of the surrounding mucosa, leading to a clear demarcation. Most elevated-type EGCs are detected easily as yellowish lesions with clearly contrasting demarcation. In some cases, a partially reddish change is accompanied on the tumor surface similar to depressed type cancer. In addition, the FICE system is quite useful for the detection of minute gastric cancer, even without magnif ication. These new contrasting images with the FICE system may have the potential to increase the rate of detection of gastric cancers and screen for them more effectively as well as to determine the extent of EGC.
文摘AIM:To conduct a preliminary study on the effect of flexible spectral imaging color enhancement (FICE) used in combination with ultraslim endoscopy by focusing on the enhanced contrast between tumor and non-tumor lesions. METHODS: We examined 50 lesions of 40 patients with epithelial tumors of the upper gastrointestinal tract before endoscopic submucosal dissection using ultraslim endoscopy with conventional natural color imag ing and with FICE imaging. We retrospectively invest igated the effect of the use of FICE on endoscopic diagn osis in comparison with normal light. RESULTS: Visibility of the epithelial tumors of the upper gastrointestinal tract with FICE was superior to normal light in 54% of the observations and comparable to normal light in 46% of the observations. There was no lesion for which visibility with FICE was inferior to that with normal light. FICE visualized 69.6% of hyperemic lesions and 58.8% of discolored lesions better than conventional endoscopy with natural color imaging. FICE sign if icantly improved the visibility of lesions with hyp ere mia or discoloration compared with normocolored lesions. CONCLUSION: This study suggests that the use of FICE would improve the ability of ultraslim endoscopy to detect epithelial tumors of the upper gastrointestinal tract.
基金Supported by Doctoral Fund of the State Education Commission of China
文摘The cross-spectral estimation methods are efficient in estimating the parameters of sinusoidal signals embedded in colored noise. But up to now, only FPT and cross-periodogram methods are used in this field, the modern auto-spectral estimation method is introduced into cross-spectral estimation in this paper, meanwhile the cross-correlation based Yule-Walker equation is proposed theoretically and the moment and singular-value decomposition (SVD)) algorithms for cross-spectral estimation have been developed. Finally, a numerical example is given for comparing the presented methods with the well-known Cadzow’s SVD method.
文摘【目的】叶面积指数(leaf area index,LAI)是表征作物长势、光合、蒸腾的重要指标。论文旨在研究不同生育期、多生育期无人机多光谱数据棉花LAI估测模型,明确不同生育期间棉花LAI估测模型变化规律,为实时掌握棉花长势并因地制宜进行田间科学管理提供依据。【方法】利用大疆精灵4多光谱无人机获取棉花现蕾期、初花期、结铃期、吐絮期多光谱图像和RGB图像。选用归一化差植被指数(NDVI)、绿度归一化差植被指数(GNDVI)、归一化差红边指数(NDRE)、叶片叶绿素指数(LCI)、优化的土壤调节植被指数(OSAVI)5种多光谱指数和修正红绿植被指数(MGRVI)、红绿植被指数(GRVI)、绿叶指数(GLA)、超红指数(EXR)、大气阻抗植被指数(VARI)5种颜色指数分别建立棉花各生育期及棉花生长多生育期数据集合,结合打孔法获取地面LAI实测数据,使用机器学习算法中偏最小二乘(PLSR)、岭回归(RR)、随机森林(RF)、支持向量机(SVM)、神经网络(BP)构建棉花LAI预测模型。【结果】覆膜棉花LAI随着生育期的变化呈现先增长后下降的趋势,现蕾期、初花期、结铃期内侧棉花叶面积指数均值均显著大于外侧(P<0.05);选择的指数在各时期彼此间均呈显著相关(P<0.05),总体而言,多光谱指数与颜色指数间的相关性随着生育期的进行而呈现下降趋势,选择的指数在各时期均与棉花LAI相关性显著(P<0.05),多光谱指数相关系数介于0.35—0.85,颜色指数相关系数介于0.49—0.71,相关系数绝对值较大的指数多为多光谱指数,颜色指数与棉花LAI的相关系数绝对值较小;估测模型性能结果显示棉花各生育期模型中多光谱指数优于颜色指数,且各指数模型预测性能随着生育期的变化呈现一定规律性,NDVI是预测棉花LAI的最优指数。从模型结果上看,RF模型和BP模型在各生育期下获得了较高的估计精度。初花期LAI反演模型精度最高,最优模型验证集R2为0.809,MAE为0.288,NRMSE为0.120。多生育期最优模型验证集R2为0.386,MAE为0.700,NRMSE为0.198。【结论】棉花内外侧LAI在现蕾期、初花期、结铃期存在显著差异。在各生育期中,RF和BP模型是预测棉花LAI较优模型。NDVI在各指数中表现最好,是预测棉花LAI的最优指数。多生育期模型效果较单生育期明显下降,最优指数为GNDVI,最优模型为BP。本研究中预测棉花LAI的最优窗口期是初花期。研究结果可为无人机遥感监测棉花LAI提供理论依据和技术支持。
文摘目的利用近红外光谱及光谱融合策略,结合化学计量学方法,建立水性油墨的颜色预测模型,实现水性油墨印刷品颜色准确预测。方法采集不同酒精含量和不同调色墨含量的油墨的近红外光谱反射率和吸光度数据,并测得对应的印刷品的Lab值,然后建立单一光谱不同预处理过后的偏最小二乘(Partial Least Squares,PLS)模型,以及基于数据层融合和特征层融合的PLS模型,最终通过比较预测集决定系数和预测集均方根误差(Root Mean Square Error of Prediction,RMSEP)以及色差来评估模型的预测效果。结果单光谱建模,基于反射率建立的模型准确率高于基于吸光度建立的模型;数据层融合缺乏稳定性,对L和b值的预测有所提升,对a值的预测几乎不变;特征层融合建模效果明显好于单一光谱和数据层融合,对Lab的预测决定系数分别达到了0.9961、0.9939、0.9974;RMSEP值分别为0.1421、0.2126、0.2072;预测值与真实值的最大色差为0.6783。结论通过光谱特征融合技术能提高油墨颜色预测精度,准确预测出油墨颜色变化。
文摘Recently,image-enhanced endoscopy(IEE) has been used to diagnose gastrointestinal tumors.This method is a change from conventional white-light(WL) endoscopy without dyeing solution,requiring only the push of a button.In IEE,there are many advantages in diagnosis of neoplastic tumors,evaluation of invasion depth for cancerous lesions,and detection of neoplastic lesions.In narrow band imaging(NBI) systems(Olympus Medical Co.,Tokyo,Japan),optical filters that allow narrow-band light to pass at wavelengths of 415 and 540 nm are used.Mucosal surface blood vessels are seen most clearly at 415 nm,which is the wavelength that corresponds to the hemoglobin absorption band,while vessels in the deep layer of the mucosa can be detected at 540 nm.Thus,NBI also can detect pit-like structures named surface pattern.The flexible spectral imaging color enhancement(FICE) system(Fujifilm Medical Co.,Tokyo,Japan) is also an IEE but different to NBI.FICE depends on the use of spectral-estimation technology to reconstruct images at different wavelengths based on WL images.FICE can enhance vascular and surface patterns.The autofluorescence imaging(AFI) video endoscope system(Olympus Medical Co.,Tokyo,Japan) is a new illumination method that uses the difference in intensity of autofluorescence between the normal area and neoplastic lesions.AFI light comprises a blue light for emitting and a green light for hemoglobin absorption.The aim of this review is to highlight the efficacy of IEE for diagnosis of colorectal tumors for endoscopic treatment.
