搜索资源列表
license-plate-recognition-system-
- 1.先打开一幅图片然后按照顺序灰度化、二值化、灰度拉伸、车牌定位、二值化、倾斜校正、字符分割、训练神经网络、识别字符。 2.测试图像存储在当前目录的img下。 3.测试集、训练集、目标向量均存储在img下的文本文件中。-First open a picture in order graying, binarization, gray stretch, license plate location, binarization, skew correction and character s
Adaboost
- Adaboost是一种迭代算法,其核心思想是针对同一个训练集训练不同的分类器(弱分类器),然后把这些弱分类器集合起来,构成一个更强的最终分类器(强分类器)。-Adaboost is an iterative algorithm, its core idea is different classifications for the same training set (weak classifiers), then these weak classifiers together to form a
pattern_recognition
- 根据训练集图像,对目标图像中的八个子目标图像进行识别,用红色矩形将各个子目标图像框起来,并在子目标图像的中央位置,写上对其识别的结果,即子目标图像的名字。-请键入文字或网站地址,或者上传文档。 取消 Gēnjù xùnliàn jí túxiàng, duì mùbiāo túxiàng zhōng de bā gè zǐ mùbiāo túxiàng jìnxíng shìbié, yòng hóngsè jǔxíng jiāng gège zǐ mùbiāo túxiàng kuān
adaboost
- Adaboost是一种迭代算法,其核心思想是针对同一个训练集训练不同的分类器(弱分类器),然后把这些弱分类器集合起来,构成一个更强的最终分类器(强分类器)。 load clouds [test_targets, E] = lijsada_boost(patterns, targets, patterns, 100, Stumps ,[]) train_patterns 每列为一样本 train_targets 每列为一样本目标 100 :Numbe
panda-face-recognition
- gui界面软件,可计算欧几里得距离识别大熊猫面部,具有大量测试集和训练集样本库-gui interface software to calculate the Euclidean distance identification panda face, with a large test set and the training set of sample library
ObjectDetection_Demo
- 这是一个针对图像跟踪的程序,可以检测人脸图像,而且不需要训练集-This is a program for video, can be very good video into a sequence of frames and preserved
pattern-recognition
- 基于特征向量的人脸识别,有训练集和样本集,通过Adaboost强分类器算法实现,结果精确度达到95 以上,给定一个example,就可以在样本集中识别出对应的人脸。-Face recognition based on feature vectors have the training set and sample set by Adaboost strong classifier algorithm, the results of more than 95 accuracy, given an
logistic-regression
- 采用随机梯度算法,来利用逻辑回归对训练集进行分类-Classify the training set by logistic regression
aam_matlab
- 主动形状模型(active appearance model)的matlab代码 训练集和测试集需要从附带的网址上下载或者根据自己的需求构造-matlab code for active appearance model, the training and test sets could be downloaded the website in readme.txt or constructed by yourself according to your problem
haarcascade_profileface
- 在opencv2.4中找到了侧脸训练集,分享一下,供大家使用-opencv: face detection
IRIS-use-RFclassification
- 用随机森林RF方法分类IRIS数据集,用一百个数据做训练集,五十个作为测试集,并统计出错误率,可直接运行-Classification method with random forests RF IRIS data set, using one hundred data to do training set, and fifty as a test set, and the statistical error rate, can be directly run
PlateIdentify
- 车牌字符识别C++程序 1.先打开一幅图片然后按照顺序灰度化、二值化、灰度拉伸、车牌定位、二值化、倾斜校正、字符分割、训练神经网络、识别字符。 2.测试图像存储在当前目录的img下。 3.测试集、训练集、目标向量均存储在img下的文本文件中。-C++ program license plate character recognition 1. Open a picture first and then follow the order of graying, binarizati
paris_sigg_release_v4.5.tar
- What makes Paris Look Like Paris 12年 sigaraph, 通过提取图像特征跟训练集对比得出地理位置-What makes Paris Look Like Paris 12 years sigaraph, by extracting image features contrast with the training set derived location
eigenface
- 用MATLAB实现人脸识别的源代码,包括训练集的实现,识别部分的书写。-Face recognition using MATLAB source code, including writing to achieve the training set, the identification section.
lpr-system
- 1.先打开一幅图片然后按照顺序灰度化、二值化、灰度拉伸、车牌定位、二值化、倾斜校正、字符分割、训练神经网络、识别字符。 2.测试图像存储在当前目录的img下。 3.测试集、训练集、目标向量均存储在img下的文本文件中。-1. Open a picture first and then according to the order gray, binarization, gray stretching, license plate localization, binarization, t
MIT_persons_jpg
- MIT行人数据库,共924张行人图片,原来是PPM格式,很多人找不到转换工具,这里已转换成JPG。该数据库只含正面和背面两个视角,无负样本,未区分训练集和测试集。-MIT pedestrian dataset of JPG format, for pedestrian detection application.
Transitive-Re-identification
- 行人再识别。人再次鉴定的准确性可以显著提高给定一个训练集,演示了外表的变化与非重叠的两个摄像头。我们测试时是否能保持这种优势直接标注的训练集并非对所有现场camera-pairs可用。给定的训练集捕捉相机A和B之间的对应关系和不同的训练集捕捉相机B和C之间的对应关系,传递鉴定算法(TRID)建议提供了一个分类器(A,C)对外观。该方法是基于统计建模和使用一个边缘化的推理过程。这种方法可以显著减少注释工作固有的学习系统。-Person re-identification accuracy can
pca
- 人脸识别,输入200副人脸照片作为训练集,检测另外200副人脸图像,计算错误率。-Recognition, input 200 people face photo as the training set, testing additional 200 people face image to calculate the error rate.
Adaboost
- Adaboost是一种迭代算法,其核心思想是针对同一个训练集训练不同的分类器(弱分类器),然后把这些弱分类器集合起来,构成一个更强的最终分类器-Adaboost is an iterative algorithm, the core idea is the same training set for different classifiers (weak classifiers), and then set up these weak classifiers to form a stronger
Adaboost
- Adaboost是一种迭代算法,其核心思想是针对同一个训练集训练不同的分类器(弱分类器),然后把这些弱分类器集合起来,构成一个更强的最终分类器-Adaboost is an iterative algorithm, the core idea is the same training set for different classifiers (weak classifiers), and then set up these weak classifiers to form a stronger