视差不连续区域和重复纹理区域的误匹配率高一直是影响双目立体匹配测量精度的主要问题,为此,本文提出一种基于多特征融合的立体匹配算法。首先,在代价计算阶段,通过高斯加权法赋予邻域像素点的权值,从而优化绝对差之和(Sum of Absolute...视差不连续区域和重复纹理区域的误匹配率高一直是影响双目立体匹配测量精度的主要问题,为此,本文提出一种基于多特征融合的立体匹配算法。首先,在代价计算阶段,通过高斯加权法赋予邻域像素点的权值,从而优化绝对差之和(Sum of Absolute Differences,SAD)算法的计算精度。接着,基于Census变换改进二进制链码方式,将邻域内像素的平均灰度值与梯度图像的灰度均值相融合,进而建立左右图像对应点的判断依据并优化其编码长度。然后,构建基于十字交叉法与改进的引导滤波器相融合的聚合方法,从而实现视差值再分配,以降低误匹配率。最后,通过赢家通吃(Winner Take All,WTA)算法获取初始视差,并采用左右一致性检测方法及亚像素法提高匹配精度,从而获取最终的视差结果。实验结果表明,在Middlebury数据集的测试中,所提SAD-Census算法的平均非遮挡区域和全部区域的误匹配率为分别为2.67%和5.69%,测量200~900 mm距离的平均误差小于2%;而实际三维测量的最大误差为1.5%。实验结果检验了所提算法的有效性和可靠性。展开更多
As a high-grade edible oil tree native in China,tea-oil tree(Camellia oleifera)has the oil-yielded rate of about 55% from its kernel.The recent researches suggested that tea-oil would be one of the best vegetable oils...As a high-grade edible oil tree native in China,tea-oil tree(Camellia oleifera)has the oil-yielded rate of about 55% from its kernel.The recent researches suggested that tea-oil would be one of the best vegetable oils,and even be better than olive oil with its abundant unsaturated fatty acids including 82.6% oleic acid.Stearoyl-ACP desaturase(SAD)is a key enzyme that catalyzes saturated fatty acids(C18∶0)bonded to ACP(Acyl carrier protein)and dehydrogenates the fatty acids into oleic acids,and hence controls the content of oleic acid and the proportion between saturated fatty acids and unsaturated fatty acids.With our previous acquisition of three cDNAs and ESTs of C.oleifera SAD(CoSAD)gene,5’RACE technology was used to obtain the full-length cDNA of CoSAD gene from the nearly matured C.oleifera seed.The comprehensive bioinformatic analyses including sequence characteristics of DNA and amino acid,multi-sequence aligning,identity and homology,molecular clustering,protein physicochemical properties,and protein structural prediction and characteristics were performed.The results may provide the theoretical and material elements for application of CoSAD gene and genetic improvement on other oil plants.展开更多
文摘视差不连续区域和重复纹理区域的误匹配率高一直是影响双目立体匹配测量精度的主要问题,为此,本文提出一种基于多特征融合的立体匹配算法。首先,在代价计算阶段,通过高斯加权法赋予邻域像素点的权值,从而优化绝对差之和(Sum of Absolute Differences,SAD)算法的计算精度。接着,基于Census变换改进二进制链码方式,将邻域内像素的平均灰度值与梯度图像的灰度均值相融合,进而建立左右图像对应点的判断依据并优化其编码长度。然后,构建基于十字交叉法与改进的引导滤波器相融合的聚合方法,从而实现视差值再分配,以降低误匹配率。最后,通过赢家通吃(Winner Take All,WTA)算法获取初始视差,并采用左右一致性检测方法及亚像素法提高匹配精度,从而获取最终的视差结果。实验结果表明,在Middlebury数据集的测试中,所提SAD-Census算法的平均非遮挡区域和全部区域的误匹配率为分别为2.67%和5.69%,测量200~900 mm距离的平均误差小于2%;而实际三维测量的最大误差为1.5%。实验结果检验了所提算法的有效性和可靠性。
文摘As a high-grade edible oil tree native in China,tea-oil tree(Camellia oleifera)has the oil-yielded rate of about 55% from its kernel.The recent researches suggested that tea-oil would be one of the best vegetable oils,and even be better than olive oil with its abundant unsaturated fatty acids including 82.6% oleic acid.Stearoyl-ACP desaturase(SAD)is a key enzyme that catalyzes saturated fatty acids(C18∶0)bonded to ACP(Acyl carrier protein)and dehydrogenates the fatty acids into oleic acids,and hence controls the content of oleic acid and the proportion between saturated fatty acids and unsaturated fatty acids.With our previous acquisition of three cDNAs and ESTs of C.oleifera SAD(CoSAD)gene,5’RACE technology was used to obtain the full-length cDNA of CoSAD gene from the nearly matured C.oleifera seed.The comprehensive bioinformatic analyses including sequence characteristics of DNA and amino acid,multi-sequence aligning,identity and homology,molecular clustering,protein physicochemical properties,and protein structural prediction and characteristics were performed.The results may provide the theoretical and material elements for application of CoSAD gene and genetic improvement on other oil plants.
