今天在GitHub上看到一个对车辆训练好的模型,即xml文件,于是拿来测试了一个效果。我用这个xml文件对视频中的每一帧画面进行简单的车辆识别定位,演示代码如下:
import cv2
import numpy as np
camera = cv2.VideoCapture ("video.avi")
camera.open("video.avi")
car_cascade = cv2.CascadeClassifier('cars.xml')
while True:
(grabbed,frame) = camera.read()
grayvideo = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
cars = car_cascade.detectMultiScale(grayvideo, 1.1, 1)
# print(cars)
# print(type(cars))
# print(cars.shape)
# 部分输出如下所示:
# [[255 62 37 37]
# [144 25 35 35]
# [219 81 62 62]
# [246 52 54 54]]
# < class 'numpy.ndarray'>
# (4, 4)
# ...
for (x,y,w,h) in cars:
cv2.rectangle(frame,(x,y),(x+w,y+h),(0,0,255),2)
cv2.imshow("video",frame)
if cv2.waitKey(1)== ord('q'):
break
camera.release()
cv2.destroyAllWindows()
运行程序,截取其中某几帧画面,如下:
显然,该模型对汽车识别的精度不够。
附:测试视频以及xml模型文件下载地址
https://download.csdn.net/download/weixin_41695564/10419268