Shulou(Shulou.com)06/02 Report--

Editor to share with you how to use python+opencv+dlib to achieve lip detection on the raspberry pie. I hope you will get something after reading this article. Let's discuss it together.

Using python+opencv+dlib to detect lips on raspberry pie

The goal of the project is to run python code on raspberry pie to achieve lip detection. It was thought that the hardware of raspberry pie could run real-time detection smoothly, but the experimental results show that real-time detection of raspberry pie is not feasible and needs to be improved later.

The results of the experiment are as follows:

1. Install related library files

The library you need to use here is opencv,numpy,dlib.

1.1 install opencv

Pip3 install opencv-python

1.2 install numpy

Numpy library is included in raspberry pie.

Pip3 install numpy

1.3 install dlib

It is relatively easy to install dlib in the raspberry pie system. All you need is pip install, but there will be errors in the window system. At this time, we need to install pip install dlib-19.17.99-cp37-cp37m-win_amd64.whl. What we need to note is that different python versions need to install the corresponding version of dlib, that is, the later "cp37-cp37m". Check the version number that can be installed by the corresponding python. You can use the command line: pip debug-- verbose to display the appropriate installation version number.

I installed cmake and dlib on the raspberry pie.

Pip3 install cmakepip3 install dlib2, code section

There are 68 points in facial features extracted by dlib.

Jaw point = 0-16

Right eyebrow point = 17-21

Left eyebrow point = 22-26

Nose point = 27-35

Right eye point = 36-41

Left eye point = 42-47

Quarrel = 48-60

Lip score = 61-67

From gpiozero import LEDfrom time import sleepfrom subprocess import check_callimport cv2import numpy as npimport dlibprint (cv2.__version__) def search_cap_num (): for i in range (2000): cap = cv2.VideoCapture (I) cap_opened = cap.isOpened () if cap_opened = = True: return I cap_num = search_cap_num () cap = cv2.VideoCapture (cap_num) detector = dlib.get_frontal _ face_detector () predictor = dlib.shape_predictor ("shape_predictor_68_face_landmarks.dat") # specifies the path of the upper lip to the lower lip lip_order_dlib = np.array ([[48] 49, 50, 51, 52, 53, 54, 64, 63, 62, 61, 60, 48], [48, 59, 58, 57, 56, 55, 54, 64, 65, 66, 67, 60, 48])-48lip_order_num = lip_order_dlib.shape [1] while 1: landmarks_lip = [] ret, frame = cap.read () gray = cv2.cvtColor (frame, cv2.COLOR_BGR2GRAY) rects = detector (gray) 1) print ('faces number:' + str (len (rects) for (I, rect) in enumerate (rects): # Mark 68 landmark points in the face landmarks = predictor (gray, rect) for n in range (48,68): X = landmarks.part (n). X y = landmarks.part (n). Y landmarks_lip.append ((x ) # cv2.circle (img=img, center= (x, y), radius=3, color= (0,255,0), thickness=-1) for m in range (lip_order_num-1): cv2.line (frame, landmarks_ lip [lip _ order_dlib [0] [m]], landmarks_ lip [lip _ order_dlib [0] [m], color= (0,255,0), thickness=2 LineType=8) for n in range (lip_order_num-1): cv2.line (frame, landmarks_ lip [lip _ order_dlib [1] [n]], landmarks_ lip [lip _ order_dlib [1] [n], color= (0,255,0), thickness=2, lineType=8) cv2.imshow ("face") Frame) if cv2.waitKey (1) & 0xff = = ord ('q'): breakcap.release () cv2.destroyAllWindows () # check_call (['sudo',' poweroff']) 3. Experimental results

Generally speaking, the effect is more stuttered, and it feels like it takes time to analyze a picture in seconds.

If only displaying the camera image is very fast, without any stutter, that is to say, if the rects = detector (gray, 1) detection command to obtain the face region does not exist in the code, then the running speed will be greatly improved. Later, you need to think about how to improve the code running speed under face detection.

After reading this article, I believe you have a certain understanding of "how to use python+opencv+dlib to detect lips on raspberry pie". If you want to know more about it, you are welcome to follow the industry information channel. Thank you for reading!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.