html5做宠物饲养网站,天津建设工程信息网网,西瓜编程网站怎么做,wordpress删除无分类文章基于深度学习的运动目标检测#xff08;三#xff09;YOLOv3识别物体 1.基于YOLOv3识别物体2.让不同类别物体的捕捉框颜色不同3.不用Matplotlib实现目标检测 目标检测#xff0c;粗略地说就是输入图片/视频#xff0c;经过处理后得到目标的位置信息#xff08;比如左上角和… 基于深度学习的运动目标检测三YOLOv3识别物体 1.基于YOLOv3识别物体2.让不同类别物体的捕捉框颜色不同3.不用Matplotlib实现目标检测 目标检测粗略地说就是输入图片/视频经过处理后得到目标的位置信息比如左上角和右下角的坐标、目标的预测类别、目标的预测置信度。前面我们阐述了不少理论知识现在需要动手实战了。对于初学者来说自己实现YOLO算法不太现实幸运的是OpenCV的DNN(Deep Neural Network)模块封装了Darknet框架封装了YOLO算法。使用OpenCV能更方便地直接运行已训练的深度学习模型本次采用在目标检测中最强劲的YOLOv3基本步骤是先让OpenCV加载预训练YOLOv3模型然后进行各种检测比如图片识别、打开计算机自带摄像头进行物体检测等。 为了加载预训练YOLOv3模型需要准备3个文件在工程目录下yolov3.cfg、yolov3.weights和coco.names。其中yolov3.cfg为yolov3网络配置文件yolov3.weights为权重文件coco.names为标签文件。
1.基于YOLOv3识别物体
使用OpenCV dnn模块加载YOLO模型代码如下 net cv2.dnn.readNet(yolov3.weights, yolov3.cfg)从coco.names导入类别并存储为列表代码如下
classes []
with open(coco.names,r)as f:classes [line.strip() for line inf.readlines()]
print(classes)完整代码
import cv2
import numpy as npnet cv2.dnn.readNet(yolov3.weights, yolov3.cfg)
classes []
with open(coco.names, r) as f: #这里使用的是coco所训练的模型yolov3.cfg所以这里对应为coco.namesclasses [line.strip() for line in f.readlines()]print(classes)layer_names net.getLayerNames()
print(layer_names)output_layers [layer_names[i[0] - 1] for i in net.getUnconnectedOutLayers()]
print(output_layers)img cv2.imread(demo1.jpg)# 获取图像尺寸与通道值
height, width, channels img.shape
print(The image height is:,height)
print(The image width is:,width)
print(The image channels is:,channels)blob cv2.dnn.blobFromImage(img, 1.0 / 255.0, (416, 416), (0, 0, 0), True, cropFalse)from matplotlib import pyplot as pltfig plt.gcf()
fig.set_size_inches(20, 10)num 0
for b in blob:for img_blob in b:img_blobcv2.cvtColor(img_blob, cv2.COLOR_BGR2RGB)num 1ax plt.subplot(3/3, 3, num)ax.imshow(img_blob)title blob_image:{}.format(num)ax.set_title(title, fontsize20)net.setInput(blob)
outs net.forward(output_layers)for i in range(len(outs)):print(The {} layer out shape is:.format(i), outs[i].shape)class_ids []
confidences []
boxes []i 0
for out in outs:for detection in out:a sum(detection[5:])if a 0:print(detection[5:])i 1if i 2:breaki 0
for out in outs:for detection in out:print(中心像素坐标 X 对原图宽比值:,detection[0])print(中心像素坐标 Y 对原图高比值:,detection[1])print(边界框的宽度 W 对原图宽比值:,detection[2])print(边界框的高度 H 对原图高比值:,detection[3])print(此边界框置信度:,detection[4])breakbreakplt_img cv2.cvtColor(img, cv2.COLOR_BGR2RGB)fig plt.gcf()
fig.set_size_inches(20, 10)plt.imshow(plt_img)# jupyter 对每次运行结果会保留再次运行列表创建
class_ids []
confidences []
boxes []i 0for out in outs:for detection in out:scores detection[5:]class_id np.argmax(scores)confidence scores[class_id]if confidence 0.5:center_x int(detection[0] * width)center_y int(detection[1] * height)w int(detection[2] * width) h int(detection[3] * height)x int(center_x - w / 2)y int(center_y - h / 2)boxes.append([x, y, w, h])confidences.append(float(confidence))class_ids.append(class_id)label classes[class_id]plt.gca().add_patch(plt.Rectangle((x, y), w,h, fillFalse,edgecolor(0, 1, 1), linewidth2))plt.text(x, y - 10, label, color (1, 0, 0), fontsize20)print(object {} :.format(i), label)i 1plt.show()plt_img cv2.cvtColor(img, cv2.COLOR_BGR2RGB)fig plt.gcf()
fig.set_size_inches(30, 20)ax_img plt.subplot(1, 2, 1)ax_img.imshow(plt_img)# jupyter 对每次运行结果会保留再次运行一次
class_ids []
confidences []
boxes []i 0for out in outs:for detection in out:scores detection[5:]class_id np.argmax(scores)confidence scores[class_id]if confidence 0.5:center_x int(detection[0] * width)center_y int(detection[1] * height)w int(detection[2] * width) h int(detection[3] * height)x int(center_x - w / 2)y int(center_y - h / 2)boxes.append([x, y, w, h])confidences.append(float(confidence))class_ids.append(class_id)label classes[class_id]plt.gca().add_patch(plt.Rectangle((x, y), w,h, fillFalse,edgecolor(0, 1, 1), linewidth2))plt.text(x, y - 10, label, color (1, 0, 0), fontsize20)print(object {} :.format(i), label *(10 - len(label)), confidence :{}.format(confidence))i 1print(confidences)
indexes cv2.dnn.NMSBoxes(boxes, confidences, 0.5, 0.4)
print(indexes, end)ax_img plt.subplot(1, 2, 2)
ax_img.imshow(plt_img)
for j in range(len(boxes)):if j in indexes:x, y, w, h boxes[j]label classes[class_ids[j]]plt.gca().add_patch(plt.Rectangle((x, y), w,h, fillFalse,edgecolor(0, 1, 1), linewidth2))plt.text(x, y - 10, label, color (1, 0, 0), fontsize20)plt.show()
获得输出层的代码 layer_names net.getLayerNames()print(layer_names)output_layers [layer_names[i[0] - 1] for i in net.getUnconnectedOutLayers()]print(output_layers)其中getLayerNames函数获取网络各层名称getUnconnectedOutLayers函数返回具有未连接输出的图层索引。
添加处理图像并获取blob的代码 img cv2.imread(demo1.jpg)# 获取图像尺寸与通道值height, width, channels img.shapeprint(The image height is:,height)print(The image width is:,width)print(The image channels is:,channels)blob cv2.dnn.blobFromImage(img, 1.0 / 255.0, (416, 416), (0, 0, 0), True,
cropFalse)此时运行程序打印的高度、宽度和通道数如下 The image height is: 2250The image width is: 4000The image channels is: 3(添加Matplotlib可视化blob下的图像代码如下 from matplotlib import pyplot as pltOpenCV采用的是BGRMatplotlib采用的是RGB需要使用cv2.COLOR_BGR2RGB将BGR转换为RGB。
利用setInput函数将blob输入网络利用forward函数输入网络输出层的名字来计算网络输出。本次计算中output_layers包含3个输出层的列表所以outs的值也是一个包含3个矩阵(array)的列表(list)。 这个循环会输出以下内容 The 0 layer out shape is: (507, 85)The 1 layer out shape is: (2028, 85)The 2 layer out shape is: (8112, 85)然后进行识别与标签处理创建记录数据列表。 其中class_ids记录类别名confidences记录算法检测物体概率boxes记录框的坐标。YOLOv3对于一个416×416的输入图像在每个尺度的特征图的每个网格中设置3个先验框总共有13×13×3 26×26×3 52×52×3 10647个预测。每一个预测是一个85(4180)维向量这个85维向量包含边框坐标4个数值、边框置信度1个数值、对象类别的概率对于COCO数据集有80种对象所以我们通过detection[5:]获取detection的后80个数据类似独热码获取其最大值索引对应的coco.names类别。 在检测中发现出现了双框或者多框效果。OpenCV dnn模块自带了NMSBoxes()函数可以使用NMS算法解决多框问题。NMS的目的是在邻域内保留同一检测目标置信度最大的框在下方输出中可以发现对于邻域相同的目标检测只保留了confidence值最大的box索引例如object 0 : tvmonitor与object 3 : tvmonitor概率分别为0.9334805607795715与0.9716598987579346显然保留了object 3 : tvmonitor在索引indexes中没有[0]元素其余推断类似。 2.让不同类别物体的捕捉框颜色不同
代码
import cv2
import numpy as np
from matplotlib import pyplot as plt# Load Yolo
net cv2.dnn.readNet(yolov3.weights, yolov3.cfg)
classes []
with open(coco.names, r) as f:classes [line.strip() for line in f.readlines()]layer_names net.getLayerNames()
output_layers [layer_names[i[0] - 1] for i in net.getUnconnectedOutLayers()]colors np.random.uniform(0, 255, size(len(classes), 3)) / 255# Loading image
img cv2.imread(demo1.jpg)
# img cv2.resize(img, None, fx0.4, fy0.4)
height, width, channels img.shape# Detecting objects
blob cv2.dnn.blobFromImage(img, 1.0 / 255.0, (416, 416), (0, 0, 0), True, cropFalse)net.setInput(blob)
outs net.forward(output_layers)# Showing informations on the screen
class_ids []
confidences []
boxes []fig plt.gcf()
fig.set_size_inches(20, 10)
plt_img cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
plt.imshow(plt_img)for out in outs:for detection in out:scores detection[5:]class_id np.argmax(scores)confidence scores[class_id]if confidence 0.5:# Object detectedcenter_x int(detection[0] * width)center_y int(detection[1] * height)w int(detection[2] * width)h int(detection[3] * height)# Rectangle coordinatesx int(center_x - w / 2)y int(center_y - h / 2)boxes.append([x, y, w, h])confidences.append(float(confidence))class_ids.append(class_id)indexes cv2.dnn.NMSBoxes(boxes, confidences, 0.5, 0.4)for i in range(len(boxes)):if i in indexes:x, y, w, h boxes[i]label str(classes[class_ids[i]])color colors[i]plt.gca().add_patch(plt.Rectangle((x, y), w,h, fillFalse,edgecolorcolor, linewidth2))plt.text(x, y - 10, label, color color, fontsize20)plt.show()
运行结果 3.不用Matplotlib实现目标检测
代码
import cv2
import numpy as np# Load Yolo
net cv2.dnn.readNet(yolov3.weights, yolov3.cfg)
classes []
with open(coco.names, r) as f:classes [line.strip() for line in f.readlines()]
layer_names net.getLayerNames()
output_layers [layer_names[i - 1] for i in net.getUnconnectedOutLayers()]
colors np.random.uniform(0, 255, size(len(classes), 3))# Loading image
img cv2.imread(demo1.jpg)
height, width, channels img.shape# Detecting objects
blob cv2.dnn.blobFromImage(img, 0.00392, (416, 416), (0, 0, 0), True, cropFalse)net.setInput(blob)
outs net.forward(output_layers)# Showing informations on the screen
class_ids []
confidences []
boxes []
for out in outs:for detection in out:scores detection[5:]class_id np.argmax(scores)confidence scores[class_id]if confidence 0.5:# Object detectedcenter_x int(detection[0] * width)center_y int(detection[1] * height)w int(detection[2] * width)h int(detection[3] * height)# Rectangle coordinatesx int(center_x - w / 2)y int(center_y - h / 2)boxes.append([x, y, w, h])confidences.append(float(confidence))class_ids.append(class_id)indexes cv2.dnn.NMSBoxes(boxes, confidences, 0.5, 0.4)font cv2.FONT_HERSHEY_SIMPLEX
for i in range(len(boxes)):if i in indexes:x, y, w, h boxes[i]label str(classes[class_ids[i]])color colors[i]cv2.rectangle(img, (x, y), (x w, y h), color, 3)cv2.putText(img, label, (x, y - 20), font, 2, color, 3)cv2.namedWindow(Image,0)
cv2.resizeWindow(Image, 1600, 900)
cv2.imshow(Image, img)
cv2.waitKey(0)
cv2.destroyAllWindows()
输出结果
