We present a novel fluorescence spectral unmixing based on target-to-background separation preprocessing, which effectively separates the multi-target fluorescence from all background autofluorescence(BF)without any h...We present a novel fluorescence spectral unmixing based on target-to-background separation preprocessing, which effectively separates the multi-target fluorescence from all background autofluorescence(BF)without any hardware-based BF acquisition and tissue specific BF estimation. Specifically, we first enhance the intrinsic accumulation contrast in target-to-background fluorescence using h-dome transformation; then separate multi-target fluorescence areas from the background in sparse multispectral data utilizing kernel maximum autocorrelation factor analysis; we further use fast marching-based image inpainting method to patch up the removed target fluorescence areas and reconstruct the multispectral BF; with the BF matrix being subtracted from the original data, the multi-target fluorophores are easily unmixed from the subtracted data using multivariate curve resolution-alternating least squares method. In two preliminary in-vivo experiments, the proposed method demonstrated excellent performance to unmix multi-target fluorescences while other state-of-art unmixing methods failed to get desired results.展开更多
The paper first discusses shortcomings of classical adjacent-frame difference. Sec ondly, based on the image energy and high order statistic(HOS) theory, background reconstruction constraints are setup. Under the help...The paper first discusses shortcomings of classical adjacent-frame difference. Sec ondly, based on the image energy and high order statistic(HOS) theory, background reconstruction constraints are setup. Under the help of block-processing technology, background is reconstructed quickly. Finally, background difference is used to detect motion regions instead of adjacent frame difference. The DSP based platform tests indicate the background can be recovered losslessly in about one second, and moving regions are not influenced by moving target speeds. The algorithm has important usage both in theory and applications.展开更多
在电力巡检过程中,无人机等边端智能检测设备往往面临输电线路绝缘子缺陷目标小、背景因素复杂等难点,且边端设备的硬件条件限制了模型的规模,导致设备算力有限,模型准确率偏低。针对上述问题,该文提出了一种基于YOLOv8-RFL(You only lo...在电力巡检过程中,无人机等边端智能检测设备往往面临输电线路绝缘子缺陷目标小、背景因素复杂等难点,且边端设备的硬件条件限制了模型的规模,导致设备算力有限,模型准确率偏低。针对上述问题,该文提出了一种基于YOLOv8-RFL(You only look once version 8-RFL)模型的输电线路绝缘子缺陷检测方法。首先,通过对原有主干网络C2f(CSPDarknet53 to 2-Stage FPN)模块进行改进,增强模型对于绝缘子缺陷的特征提取能力;其次,构建基于特征聚焦的泛化特征金字塔网络(focusing generalized feature pyramid networks,FGFPN),采用“特征聚焦-扩散”的思想,精细化小缺陷目标的特征表达;然后,设计基于交叉注意机制的特征语义融合模块(feature semantic fusion module,FSFM),优化了对关键特征信息的捕获和利用;最后,提出轻量化权重共享检测头(Lightweight weight sharing detection head,LWSD),在保证检测精度的同时提高模型的计算效率和实时性。实验表明,改进后的YOLOv8-RFL模型均值平均精度(mean average precision,mAP)达到了93.2%,相较于基准模型提升了5.9%,在降低模型参数量和所需计算量的同时,实现了更好的绝缘子小目标缺陷检测效果,对于复杂背景下的输电线路绝缘子缺陷检测具有一定的现实意义。展开更多
基金the Small Animal Imaging Project supported by Geneway Biotech International Trading Co.,Ltd.(No.06-545)the National Natural Science Foundation of China(Nos.61271320,60872102 and 60402021)
文摘We present a novel fluorescence spectral unmixing based on target-to-background separation preprocessing, which effectively separates the multi-target fluorescence from all background autofluorescence(BF)without any hardware-based BF acquisition and tissue specific BF estimation. Specifically, we first enhance the intrinsic accumulation contrast in target-to-background fluorescence using h-dome transformation; then separate multi-target fluorescence areas from the background in sparse multispectral data utilizing kernel maximum autocorrelation factor analysis; we further use fast marching-based image inpainting method to patch up the removed target fluorescence areas and reconstruct the multispectral BF; with the BF matrix being subtracted from the original data, the multi-target fluorophores are easily unmixed from the subtracted data using multivariate curve resolution-alternating least squares method. In two preliminary in-vivo experiments, the proposed method demonstrated excellent performance to unmix multi-target fluorescences while other state-of-art unmixing methods failed to get desired results.
文摘The paper first discusses shortcomings of classical adjacent-frame difference. Sec ondly, based on the image energy and high order statistic(HOS) theory, background reconstruction constraints are setup. Under the help of block-processing technology, background is reconstructed quickly. Finally, background difference is used to detect motion regions instead of adjacent frame difference. The DSP based platform tests indicate the background can be recovered losslessly in about one second, and moving regions are not influenced by moving target speeds. The algorithm has important usage both in theory and applications.
文摘在电力巡检过程中,无人机等边端智能检测设备往往面临输电线路绝缘子缺陷目标小、背景因素复杂等难点,且边端设备的硬件条件限制了模型的规模,导致设备算力有限,模型准确率偏低。针对上述问题,该文提出了一种基于YOLOv8-RFL(You only look once version 8-RFL)模型的输电线路绝缘子缺陷检测方法。首先,通过对原有主干网络C2f(CSPDarknet53 to 2-Stage FPN)模块进行改进,增强模型对于绝缘子缺陷的特征提取能力;其次,构建基于特征聚焦的泛化特征金字塔网络(focusing generalized feature pyramid networks,FGFPN),采用“特征聚焦-扩散”的思想,精细化小缺陷目标的特征表达;然后,设计基于交叉注意机制的特征语义融合模块(feature semantic fusion module,FSFM),优化了对关键特征信息的捕获和利用;最后,提出轻量化权重共享检测头(Lightweight weight sharing detection head,LWSD),在保证检测精度的同时提高模型的计算效率和实时性。实验表明,改进后的YOLOv8-RFL模型均值平均精度(mean average precision,mAP)达到了93.2%,相较于基准模型提升了5.9%,在降低模型参数量和所需计算量的同时,实现了更好的绝缘子小目标缺陷检测效果,对于复杂背景下的输电线路绝缘子缺陷检测具有一定的现实意义。
