/// <summary>
/// 用模板去检测图像中的物体
/// </summary>
/// <param name="bmpSource">源图像，需24位或32位真彩位图</param>
/// <param name="template">基础分类器模板</param>
/// <param name="fern">Fern</param>
/// <returns></returns>
public static RectangleCollection DetectObject(Bitmap bmpSource, BaseClassifierTemplate template, Fern fern)
{
RectangleCollection rc = new RectangleCollection();
Bitmap bmp = bmpSource;
if (bmpSource.Width < template.BmpWidth || bmpSource.Height < template.BmpHeight)
{
return(null);
}
if (bmp != null && (bmp.PixelFormat == PixelFormat.Format24bppRgb || bmp.PixelFormat ==
PixelFormat.Format32bppRgb || bmp.PixelFormat == PixelFormat.Format32bppArgb))
{
UInt32[] featurecode = null;
bmp = ImgOper.Grayscale(bmpSource);
int[,] igram = ImgOper.Integrogram(bmp, 1);
while (template.BmpWidth < bmp.Width && template.BmpHeight < bmp.Height)
{
for (int y = 0; y < bmp.Height - template.BmpHeight + 1; y += (template.BmpHeight / 10))
{
for (int x = 0; x < bmp.Width - template.BmpWidth + 1; x += (template.BmpWidth / 10))
{
//int posnum = 0;
//int negnum = 0;
featurecode = BaseClassifierTemplate.GetFeatureCodeGroup(igram, bmp.Width, bmp.Height, template, x, y);
double prob = 0;
for (int i = 0; i < template.GroupNum; i++)
{
prob += fern.Probability[featurecode[i]];
//prob = fern.Probability[featurecode[i]];
//if (prob > 0.5)
//{
// posnum++;
//}
//else
//{
// negnum++;
//}
}
prob = prob / template.GroupNum;
if (prob > 0.5)
//if (posnum > negnum)
{
Rectangle rect = new Rectangle(x, y, template.BmpWidth, template.BmpHeight);
rc.Add(rect);
}
}
}
template.ResizeTemplate(1.2);
}
}
return(rc);
}
/// <summary>
/// 负样本检测专家，专门检测被误判为正的负样本和增加负样本集合
/// </summary>
/// <param name="detectCollection">检测模块产生的区域集合</param>
/// <param name="trackerRect">跟踪模块产生的区域</param>
/// <param name="bmp">被检测的位图</param>
/// <returns>返回最可信对象区域</returns>
public Rectangle NegativeExpert(RectangleCollection detectCollection, RectangleF trackerRect, Bitmap bmp)
{
// 复制一个矩形集合是为了让NExpert独立与PExpert
RectangleCollection newRectCollection = new RectangleCollection();
if (detectCollection != null)
{
foreach (Rectangle detectrect in detectCollection)
{
newRectCollection.Add(detectrect);
}
}
if (trackerRect != Rectangle.Empty)
{
// 将跟踪到的目标也加入待评估的对象集合
newRectCollection.Add(new Rectangle((int)trackerRect.X, (int)trackerRect.Y, (int)trackerRect.Width, (int)trackerRect.Height));
}
DateTime dt = DateTime.Now;
// 最可信的对象
Rectangle confidentRect = MinDistanceObject(newRectCollection, bmp);
double elapse = DateTime.Now.Subtract(dt).TotalMilliseconds;
if (confidentRect != Rectangle.Empty)
{
newRectCollection.Remove(confidentRect);
}
dt = DateTime.Now;
foreach (Rectangle rect in newRectCollection)
{
// 判断目标是否确实为背景，正距离归一化系数大于0.5
// 目标加入负样本队列的第一道关，目标必须看起来像负样本（正距离归一化系数大于0.5）
if (MinDistance(rect, bmp) > Parameter.MEDIAN_COEF)
{
double areainsect = AreaProportion(confidentRect, rect);
// 与最可信对象交集面积小于AREA_INTERSECT_PROPORTION的为负样本，加入负样本列表
if (areainsect < Parameter.AREA_INTERSECT_PROPORTION)
{
Bitmap patch = ImgOper.CutImage(bmp, rect.X, rect.Y, rect.Width, rect.Height);
TrainNegative(patch);
}
}
}
elapse = DateTime.Now.Subtract(dt).TotalMilliseconds;
return(confidentRect);
}
/// <summary>
/// 正样本检测专家，专门检测被误判为负的正样本
/// </summary>
/// <param name="detectCollection">检测模块产生的区域集合</param>
/// <param name="trackerRect">跟踪模块产生的区域</param>
/// <param name="bmp">被检测位图</param>
public void PositiveExpert(RectangleCollection detectCollection, RectangleF trackerRect, Bitmap bmp)
{
double areaproportion = 0;
bool nointersect = true; // 指明跟踪模块和检测模块得到的区域是否无交集
if (trackerRect == Rectangle.Empty)
{
return;
}
foreach (Rectangle rect in detectCollection)
{
areaproportion = AreaProportion(trackerRect, rect);
if (areaproportion > Parameter.AREA_INTERSECT_PROPORTION)
{
nointersect = false;
break;
}
}
// 没有交集，说明存在被误判为负的正样例
if (nointersect)
{
// 判断跟踪到的目标是否确实为要识别的物体，正距离归一化系数小于0.5
// 目标加入正样本队列的第一道关，目标必须看起来像正样本（正距离归一化系数小于0.5）
if (MinDistance(trackerRect, bmp) < Parameter.MEDIAN_COEF)
{
for (double lrshift = (-1) * Parameter.SHIFT_BORDER; lrshift < Parameter.SHIFT_BORDER + Parameter.SHIFT_INTERVAL; lrshift += Parameter.SHIFT_INTERVAL)
{
for (double tbshift = (-1) * Parameter.SHIFT_BORDER; tbshift < Parameter.SHIFT_BORDER + Parameter.SHIFT_INTERVAL; tbshift += Parameter.SHIFT_INTERVAL)
{
if (trackerRect.X + lrshift >= 0 && trackerRect.X + trackerRect.Width - 1 + lrshift < bmp.Width - 1 &&
trackerRect.Y + tbshift >= 0 && trackerRect.Y + trackerRect.Height - 1 + tbshift < bmp.Height - 1)
{
Bitmap patch = ImgOper.CutImage(bmp, (int)(trackerRect.X + lrshift), (int)(trackerRect.Y + tbshift),
(int)trackerRect.Width, (int)trackerRect.Height);
TrainPositive(patch);
}
}
}
}
}
}
/// <summary>
/// 从目标集合中找出最大相关系数的目标，即最可信对象
/// </summary>
/// <param name="rectCollection">目标集合</param>
/// <param name="bmp">位图</param>
/// <returns>最可信对象</returns>
public Rectangle MostAssociateObject(RectangleCollection rectCollection, Bitmap bmp)
{
double coef = 0;
double maxcoef = 0;
Rectangle confidentRect = Rectangle.Empty;
if (rectCollection == null || rectCollection.Count == 0)
{
return(confidentRect);
}
foreach (Rectangle rect in rectCollection)
{
coef = MostAssociate(rect, bmp);
if (maxcoef < coef)
{
maxcoef = coef;
confidentRect = rect;
}
}
return(confidentRect);
}
/// <summary>
/// 在疑似目标集合中找到最可信的目标
/// </summary>
/// <param name="rectCollection">目标集合</param>
/// <param name="bmp">位图</param>
/// <returns>最可信目标</returns>
private Rectangle MostConfidentObject(RectangleCollection rectCollection, Bitmap bmp)
{
double distcoef = double.MaxValue;
double mindistcoef = double.MaxValue;
Rectangle confidentRect = Rectangle.Empty;
if (rectCollection == null || rectCollection.Count == 0)
{
return(confidentRect);
}
foreach (Rectangle rect in rectCollection)
{
distcoef = NearestNeighbour(rect, bmp);
if (distcoef < mindistcoef)
{
mindistcoef = distcoef;
confidentRect = rect;
}
}
return(confidentRect);
}
private void videoSourcePlayer1_NewFrame(object sender, ref Bitmap image)
{
Bitmap nowImg = AForge.Imaging.Image.Clone(image);
nowImg = ImgOper.Grayscale(nowImg);
// 做高斯模糊，取样和检测都会影响到
nowImg = ImgOper.GaussianConvolution(nowImg, Parameter.GAUSSIAN_SIGMA, Parameter.GAUSSIAN_SIZE);
// 将图像传递到取样窗口
pri_bmp = AForge.Imaging.Image.Clone(nowImg);
double elapse = 0;
DateTime dt = DateTime.Now;
pri_obj_regions = pri_tld.HogDetect(nowImg);
elapse = DateTime.Now.Subtract(dt).TotalMilliseconds;
float[] vect = new float[3]; // 描述区域位移和缩放
//nowImg.Save("Image\\VideoSave\\" + DateTime.Now.ToString("yyyyMMddhhmmss") + ".jpg");
ArrayList points = null;
if (pri_optical.IPoints == null || pri_tracker_rect == Rectangle.Empty)
{
pri_optical.I = pri_optical.TransformBmpToLayerImg(nowImg);
// 尚未确定跟踪对象，不需要PositiveExpert
pri_tracker_rect = pri_tld.NegativeExpert(pri_obj_regions, pri_tracker_rect, nowImg);
pri_optical.IPoints = new ArrayList();
points = pri_optical.ChooseRectRandomPoints(pri_tracker_rect, Parameter.INITIAL_POINTS_NUMBER);
foreach (PointF pt in points)
{
pri_optical.IPoints.Add(new PointF(pt.X, pt.Y));
}
}
else
{
if (pri_optical.J != null)
{
pri_optical.I = pri_optical.J;
pri_optical.IPoints = pri_optical.JPoints;
}
pri_optical.J = pri_optical.TransformBmpToLayerImg(nowImg);
//pri_optical.JPoints = new ArrayList();
// 跟踪时wx和wy设为3，或者更大
dt = DateTime.Now;
vect = pri_optical.ComputerDisplacement(pri_optical.I, pri_optical.J, pri_optical.IPoints,
1, 3, 3, 20, 0.2f, ref pri_optical.IPoints, ref pri_optical.JPoints);
elapse = DateTime.Now.Subtract(dt).TotalMilliseconds;
if (vect != null)
{
pri_tracker_rect.X = pri_tracker_rect.X + vect[0];
pri_tracker_rect.Y = pri_tracker_rect.Y + vect[1];
pri_tracker_rect.Width = pri_tracker_rect.Width * vect[2];
pri_tracker_rect.Height = pri_tracker_rect.Height * vect[2];
if (pri_tracker_rect.X < 0)
{
pri_tracker_rect.Width = pri_tracker_rect.Width + pri_tracker_rect.X;
pri_tracker_rect.X = 0;
}
else if (pri_tracker_rect.X + pri_tracker_rect.Width - 1 > nowImg.Width - 1)
{
pri_tracker_rect.Width = nowImg.Width - pri_tracker_rect.X;
}
if (pri_tracker_rect.Y < 0)
{
pri_tracker_rect.Height = pri_tracker_rect.Height + pri_tracker_rect.Y;
pri_tracker_rect.Y = 0;
}
else if (pri_tracker_rect.Y + pri_tracker_rect.Height - 1 > nowImg.Height - 1)
{
pri_tracker_rect.Height = nowImg.Height - pri_tracker_rect.Y;
}
}
else
{
pri_tracker_rect = Rectangle.Empty;
}
// 在跟踪框被NExpert产生的最可信对象替代前保存原始状态
if (pri_tracker_rect != Rectangle.Empty)
{
pri_obj_rect = new RectangleF(pri_tracker_rect.X, pri_tracker_rect.Y, pri_tracker_rect.Width, pri_tracker_rect.Height);
}
dt = DateTime.Now;
// 有跟踪对象时，PositvieExpert与NegativeExpert都开始工作
pri_tld.PositiveExpert(pri_obj_regions, pri_tracker_rect, nowImg);
elapse = DateTime.Now.Subtract(dt).TotalMilliseconds;
dt = DateTime.Now;
pri_tracker_rect = pri_tld.NegativeExpert(pri_obj_regions, pri_tracker_rect, nowImg);
elapse = DateTime.Now.Subtract(dt).TotalMilliseconds;
pri_optical.JPoints = new ArrayList();
points = pri_optical.ChooseRectRandomPoints(pri_tracker_rect, Parameter.INITIAL_POINTS_NUMBER);
// 如果pri_tracker_rect为空，则pri_optical.JPoints为空，那么下次跟踪的pri_optical.Ipoints也为空，无法继续跟踪
foreach (PointF pt in points)
{
pri_optical.JPoints.Add(new PointF(pt.X, pt.Y));
}
}
Graphics g = Graphics.FromImage(image);
Pen pen = new Pen(Color.Red);
if (pri_tracker_rect != Rectangle.Empty)
{
g.DrawRectangle(pen, pri_tracker_rect.X, pri_tracker_rect.Y, pri_tracker_rect.Width, pri_tracker_rect.Height);
FontFamily f = new FontFamily("宋体");
Font font = new System.Drawing.Font(f, 12);
SolidBrush myBrush = new SolidBrush(Color.Blue);
StringBuilder str = new StringBuilder();
if (vect != null)
{
str.AppendFormat("跟踪框相对位移：({0}, {1})，缩放：{2} \r\n" +
"跟踪产生的Rectangle：({3}, {4}, {5}, {6}) \r\n" +
"最可信的Rectangle：({7}, {8}, {9}, {10})", vect[0], vect[1], vect[2],
pri_obj_rect.X, pri_obj_rect.Y, pri_obj_rect.Width, pri_obj_rect.Height,
pri_tracker_rect.X, pri_tracker_rect.Y, pri_tracker_rect.Width, pri_tracker_rect.Height);
}
else
{
str.AppendFormat("最可信的Rectangle：({0}, {1}, {2}, {3})", pri_tracker_rect.X, pri_tracker_rect.Y, pri_tracker_rect.Width, pri_tracker_rect.Height);
}
g.DrawString(str.ToString(), font, myBrush, 0, 0);
}
//if (pri_obj_regions != null && pri_obj_regions.Count > 0)
//{
// foreach (Rectangle rect in pri_obj_regions)
// {
// g.DrawRectangle(pen, rect.X, rect.Y, rect.Width, rect.Height);
// }
//}
}
/// <summary>
/// Hog检测, 被检测到图像自动缩放到BMPLIMITSIZE容忍范围内，并在检测完后将检测框自动放大之前缩小的倍率
/// </summary>
/// <param name="bmp">位图</param>
public RectangleCollection HogDetect(Bitmap bmp)
{
RectangleCollection resultCollection = null;
if (bmp == null)
{
return(null);
}
if (NegCenter == null && PosCenter == null)
{
return(null);
}
DateTime dt = DateTime.Now;
double elapse = 0;
// 针对原图的缩放倍率
double se = 1;
if (bmp.Width > Parameter.BMPLIMITSIZE.Width || bmp.Height > Parameter.BMPLIMITSIZE.Height)
{
se = bmp.Width / (double)Parameter.BMPLIMITSIZE.Width > bmp.Height / (double)Parameter.BMPLIMITSIZE.Height ?
bmp.Width / (double)Parameter.BMPLIMITSIZE.Width : bmp.Height / (double)Parameter.BMPLIMITSIZE.Height;
bmp = ImgOper.ResizeImage(bmp, (int)(bmp.Width / se), (int)(bmp.Height / se));
}
bmp = ImgOper.Grayscale(bmp);
//bmp = ImgOper.GaussianConvolution(bmp, GAUSSIAN_SIGMA, GAUSSIAN_SIZE); // 高斯卷积，使得图像平滑
// 所有层的检测结果
ArrayList resultlayers = new ArrayList();
// 初始缩放因子
double scalecoef = 1.0;
Bitmap scalebmp = null;
int newwidth = (int)(bmp.Width / scalecoef);
int newheight = (int)(bmp.Height / scalecoef);
// 每层最小距离点的集合
ArrayList idx_layermindistance = new ArrayList();
int cnt = 0;
do
{
scalebmp = ImgOper.ResizeImage(bmp, newwidth, newheight);
HogGram hogGram = HogGram.GetHogFromBitmap(scalebmp, Parameter.CELL_SIZE.Width, Parameter.CELL_SIZE.Height, Parameter.PART_NUMBER);
NormBlockVectorGram blockGram = new NormBlockVectorGram(hogGram, Parameter.BLOCK_SIZE.Width, Parameter.BLOCK_SIZE.Height);
DetectResultLayer detectlayer = new DetectResultLayer();
// !!!!!!检测窗口的像素尺寸必须能被cell尺寸整除!!!!!!像素尺寸除以hog尺寸就是检测窗口的尺寸
detectlayer.DetectResult = blockGram.DetectImgByHogWindow(
new Size(Parameter.DETECT_WINDOW_SIZE.Width / Parameter.CELL_SIZE.Width, Parameter.DETECT_WINDOW_SIZE.Height / Parameter.CELL_SIZE.Height),
NegCenter, PosCenter, Parameter.POS_DIST_COEF);
if (detectlayer.DetectResult == null)
{
return(null);
}
detectlayer.ScaleCoef = scalecoef;
resultlayers.Add(detectlayer); // 本层检测结果加入队列
scalecoef *= Parameter.SCALE_COEF; // 逐次缩小图像
newwidth = (int)(bmp.Width / scalecoef);
newheight = (int)(bmp.Height / scalecoef);
cnt++;
} while (newwidth > 2 * Parameter.DETECT_WINDOW_SIZE.Width && newheight > 2 * Parameter.DETECT_WINDOW_SIZE.Height);
elapse = DateTime.Now.Subtract(dt).TotalSeconds;
// 框出所有可能的物体
WindowResult[] wr = null;
Rectangle rect;
double mindist = -1;
WindowResult min_obj = null;
double min_scalecoef = 1;
resultCollection = new RectangleCollection();
foreach (DetectResultLayer layer in resultlayers)
{
wr = layer.DetectResult;
for (int i = 0; i < wr.Length; i++)
{
if (wr[i].label == 1)
{
if (mindist == -1 || mindist > wr[i].PosDistance)
{
mindist = wr[i].PosDistance;
min_obj = wr[i];
min_scalecoef = layer.ScaleCoef;
}
rect = new Rectangle((int)(wr[i].ImageRegion.X * layer.ScaleCoef * se),
(int)(wr[i].ImageRegion.Y * layer.ScaleCoef * se),
(int)(wr[i].ImageRegion.Width * layer.ScaleCoef * se),
(int)(wr[i].ImageRegion.Height * layer.ScaleCoef * se));
resultCollection.Add(rect);
}
}
}
//rect = new Rectangle((int)(min_obj.ImageRegion.X * min_scalecoef * se),
// (int)(min_obj.ImageRegion.Y * min_scalecoef * se),
// (int)(min_obj.ImageRegion.Width * min_scalecoef * se),
// (int)(min_obj.ImageRegion.Height * min_scalecoef * se));
//resultCollection.Add(rect);
return(resultCollection);
}