boost.h

最新推荐文章于 2023-05-01 08:00:00 发布

米粒的光芒

最新推荐文章于 2023-05-01 08:00:00 发布

阅读量695

点赞数

分类专栏： OpenCV

本文链接：https://blog.csdn.net/wobujianni/article/details/50017969

版权

OpenCV 专栏收录该内容

1 篇文章 0 订阅

订阅专栏

class CvCascadeBoost : public CvBoost
{
public:
    virtual bool train( const CvFeatureEvaluator* _featureEvaluator,
                        int _numSamples, int _precalcValBufSize, int _precalcIdxBufSize,
                        const CvCascadeBoostParams& _params=CvCascadeBoostParams() );
    virtual float predict( int sampleIdx, bool returnSum = false ) const;

    float getThreshold() const { return threshold; }
    void write( cv::FileStorage &fs, const cv::Mat& featureMap ) const;
    bool read( const cv::FileNode &node, const CvFeatureEvaluator* _featureEvaluator,
               const CvCascadeBoostParams& _params );
    void markUsedFeaturesInMap( cv::Mat& featureMap );
protected:
    virtual bool set_params( const CvBoostParams& _params );
    virtual void update_weights( CvBoostTree* tree );
    virtual bool isErrDesired();

    float threshold;
    float minHitRate, maxFalseAlarm;
};

//开始训练强分类器。输入分别为：特征向量、正负样本总数、特征值所占内存量，索引所占内存量，参数
bool CvCascadeBoost::train( const CvFeatureEvaluator* _featureEvaluator,
int _numSamples,
int _precalcValBufSize, int _precalcIdxBufSize,
const CvCascadeBoostParams& _params )
{
bool isTrained = false;
CV_Assert( !data );
clear();//清理本对象中的矩阵所占内存等。
//读取数据
data = new CvCascadeBoostTrainData( _featureEvaluator, _numSamples,
_precalcValBufSize, _precalcIdxBufSize, _params );
//弱分类器所需内存
CvMemStorage *storage = cvCreateMemStorage();
weak = cvCreateSeq( 0, sizeof(CvSeq), sizeof(CvBoostTree*), storage );
storage = 0;

//设置参数
set_params( _params );
if ( (_params.boost_type == LOGIT) || (_params.boost_type == GENTLE) )
    data->do_responses_copy();
//更新权重
update_weights( 0 );

cout << "+----+---------+---------+" << endl;
cout << "|  N |    HR   |    FA   |" << endl;
cout << "+----+---------+---------+" << endl;

do
{
    CvCascadeBoostTree* tree = new CvCascadeBoostTree;//弱分类器
    if( !tree->train( data, subsample_mask, this ) )//训练
    {
        delete tree;
        break;
    }
    cvSeqPush( weak, &tree );//保存弱分类器
    update_weights( tree );//更新权重
    trim_weights();
    if( cvCountNonZero(subsample_mask) == 0 )
        break;
}
while( !isErrDesired() && (weak->total < params.weak_count) );

if(weak->total > 0)
{
    data->is_classifier = true;
    data->free_train_data();
    isTrained = true;
}
else
    clear();

return isTrained;

}