对二值水印图像预处理后，根据分类结果采用不同强度自适应的嵌入水印，嵌入采用奇偶判决法则。
Then after preprocessing the binary image watermark, with different embedding strength the watermark image is embedded by the rule of odd-even adjudgement.
该算法利用背景差法获得前景图像，然后进行二值化和形态学处理，再和背景帧进行比较来对滞留和搬移物体进行检测和分类。
This new algorithm USES the background subtraction method to get the foreground image, and then gets the binary image using threshold value and makes the morphology processing.
矩不变量在图像进行平移、旋转和尺度变换时保持定值，因此被广泛使用在目标识别、图像分类、图像压缩和场景匹配等各种领域。
The invariable moment, which is invariable to the translation, scaling and rotation of image, is widely used in object recognition, image classification, image condensation, scene matching and so on.
文中采用模糊识别的绝对值指数法对高光谱图像进行识别分类，并对分类结果进行像元级的评价。
The article used absolute exponent as one of the methods of fuzzy diagnosis to identify and classify hyperspectral images, then evaluated precision of classification result based on pixel level.
首先基于差分图像和肤色信息检测出人脸，其次使用改进的奇异值分解方法提取面部特征，最后运用最小距离分类器进行识别。
First based on the difference image and the skin color the face is detected and localized, then features are extracted by using improved SVD, last the features is classified by minimal distance.
对二值水印图像预处理后，根据分类结果自适应的嵌入水印，嵌入采用奇偶判决法则。
Then after preprocessing the binary image watermark, with different embedding strength the watermark image is embedded by the rule of odd-even adjudging.
本文提出一种复杂性度量法对肝脏B超图像纹理的分类进行了研究，图像的灰度值复杂性作为纹理特征。
This paper presents a method for analyzing the ultrasonic liver images textures. The texture signature is taken to be a complexity of image gray levels.
子图像的特征奇异值组成整个图像的局部奇异值向量，作为分类器的输入。
With the data of characteristic singular values of sub-images, the Local Singular Value Vector of the whole image are combined, and are used as the input of the classifier.
将经过预处理的医学图像提取特征值，然后利用基于改进的小波神经网络算法的分类器对医学图像进行分类。
It extracted eigenvalues from pretreated medical images, and then classified medical images by using classifier based on improved wavelet neural network algorithm.
这种算法基于二值图像，先从二值图像中找出各数字的特征，然后再用决策树对这些数字进行分类。
The method is based on a binary image. First, the authors extract the features of every numeral in the image, then classify the numerals by a decision tree.
然后对人脸图像做奇异值分解和离散傅立叶变换，并分别提取最佳鉴别变换特征，用最近邻方法进行分类。
The optimal discriminant features of face are extracted using singular value decomposition and Fourier transform, and then they are classified by the nearest neighbor method. In...
由于文本连通分量和非文本连通分量在特征上存在差异，大多数非文本会被级联分类器丢弃，而SVM则能在此结果上做进一步的验证，因此最终输出只有文本的二值图像。
Most of non-text CCs are filtered out by cascade classifier and the remaining CCs are further verified by SVM. The final outputs are binary images containing texts only.
由于文本连通分量和非文本连通分量在特征上存在差异，大多数非文本会被级联分类器丢弃，而SVM则能在此结果上做进一步的验证，因此最终输出只有文本的二值图像。
Most of non-text CCs are filtered out by cascade classifier and the remaining CCs are further verified by SVM. The final outputs are binary images containing texts only.
应用推荐