一、簡介git
本項目的目的是爲了給你們提供跟多的實戰思路，拋磚引玉爲你們提供一個案例，也但願讀者能夠根據該方法實現更多的思想與想法，也但願讀者能夠改進該項目種提到的方法，好比改進其中的行人檢測器、跟蹤方法、行爲識別算法等等。
github
本項目主要檢測識別的行爲有7類：Standing, Walking, Sitting, Lying Down, Stand up, Sit down, Fall Down。web
二、項目方法簡介
本文涉及的方法與算法包括：YOLO V3 Tiny、Deepsort、ST-GCN方法，其中YOLO V3 Tiny用於行人檢測、DeepSort用於跟蹤、而ST-GCN則是用於行爲檢測。
算法
這裏因爲YOLO與DeepSort你們都已經比較瞭解，所以這裏只簡單說明一下ST-GCN 的流程，這裏ST-GCN 的方法結構圖以下：微信
給出一個動做視頻的骨架序列信息，首先構造出表示該骨架序列信息的圖結構，ST-GCN的輸入就是圖節點上的關節座標向量，而後是一系列時空圖卷積操做來提取高層的特徵，最後用SofMax分類器獲得對應的動做分類。整個過程實現了端到端的訓練。網絡
GCN 幫助咱們學習了到空間中相鄰關節的局部特徵。在此基礎上，咱們須要學習時間中關節變化的局部特徵。如何爲 Graph 疊加時序特徵，是圖卷積網絡面臨的問題之一。這方面的研究主要有兩個思路：時間卷積（TCN）和序列模型（LSTM）。app
ST-GCN 使用的是 TCN，因爲形狀固定，可使用傳統的卷積層完成時間卷積操做。爲了便於理解，能夠類比圖像的卷積操做。st-gcn 的 feature map 最後三個維度的形狀爲(C,V,T)，與圖像 feature map 的形狀(C,W,H)相對應。編輯器
圖像的通道數C對應關節的特徵數C。ide
圖像的寬W對應關鍵幀數V。函數
圖像的高H對應關節數T。
在圖像卷積中，卷積核的大小爲『w』×『1』，則每次完成w行像素，1列像素的卷積。『stride』爲s，則每次移動s像素，完成1行後進行下1行像素的卷積。
在時間卷積中，卷積核的大小爲『temporal_kernel_size』×『1』，則每次完成1個節點，temporal_kernel_size 個關鍵幀的卷積。『stride』爲1，則每次移動1幀，完成1個節點後進行下1個節點的卷積。
訓練以下：
輸入的數據首先進行batch normalization，而後在通過9個ST-GCN單元，接着是一個global pooling獲得每一個序列的256維特徵向量，最後用SoftMax函數進行分類，獲得最後的標籤。
每個ST-GCN採用Resnet的結構，前三層的輸出有64個通道，中間三層有128個通道，最後三層有256個通道，在每次通過ST-CGN結構後，以0.5的機率隨機將特徵dropout，第4和第7個時域卷積層的strides設置爲2。用SGD訓練，學習率爲0.01，每10個epochs學習率降低0.1。
ST-GCN 最末卷積層的響應可視化結果圖以下：
本文項目主函數代碼以下：
import os
import cv2
import time
import torch
import argparse
import numpy as np
from Detection.Utils import ResizePadding
from CameraLoader import CamLoader, CamLoader_Q
from DetectorLoader import TinyYOLOv3_onecls
from PoseEstimateLoader import SPPE_FastPose
from fn import draw_single
from Track.Tracker import Detection, Tracker
from ActionsEstLoader import TSSTG
# source = '../Data/test_video/test7.mp4'
# source = '../Data/falldata/Home/Videos/video (2).avi' # hard detect
source = './output/test3.mp4'
# source = 2
def preproc(image):
"""preprocess function for CameraLoader.
"""
image = resize_fn(image)
image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
return image
def kpt2bbox(kpt, ex=20):
"""Get bbox that hold on all of the keypoints (x,y)
kpt: array of shape `(N, 2)`,
ex: (int) expand bounding box,
"""
return np.array((kpt[:, 0].min() - ex, kpt[:, 1].min() - ex,
kpt[:, 0].max() + ex, kpt[:, 1].max() + ex))
if __name__ == '__main__':
par = argparse.ArgumentParser(description='Human Fall Detection Demo.')
par.add_argument('-C', '--camera', default=source, # required=True, # default=2,
help='Source of camera or video file path.')
par.add_argument('--detection_input_size', type=int, default=384,
help='Size of input in detection model in square must be divisible by 32 (int).')
par.add_argument('--pose_input_size', type=str, default='224x160',
help='Size of input in pose model must be divisible by 32 (h, w)')
par.add_argument('--pose_backbone', type=str, default='resnet50', help='Backbone model for SPPE FastPose model.')
par.add_argument('--show_detected', default=False, action='store_true', help='Show all bounding box from detection.')
par.add_argument('--show_skeleton', default=True, action='store_true', help='Show skeleton pose.')
par.add_argument('--save_out', type=str, default='./output/output3.mp4', help='Save display to video file.')
par.add_argument('--device', type=str, default='cuda', help='Device to run model on cpu or cuda.')
args = par.parse_args()
device = args.device
# DETECTION MODEL.
inp_dets = args.detection_input_size
detect_model = TinyYOLOv3_onecls(inp_dets, device=device)
# POSE MODEL.
inp_pose = args.pose_input_size.split('x')
inp_pose = (int(inp_pose[0]), int(inp_pose[1]))
pose_model = SPPE_FastPose(args.pose_backbone, inp_pose[0], inp_pose[1], device=device)
# Tracker.
max_age = 30
tracker = Tracker(max_age=max_age, n_init=3)
# Actions Estimate.
action_model = TSSTG()
resize_fn = ResizePadding(inp_dets, inp_dets)
cam_source = args.camera
if type(cam_source) is str and os.path.isfile(cam_source):
# Use loader thread with Q for video file.
cam = CamLoader_Q(cam_source, queue_size=1000, preprocess=preproc).start()
else:
# Use normal thread loader for webcam.
cam = CamLoader(int(cam_source) if cam_source.isdigit() else cam_source,
preprocess=preproc).start()
# frame_size = cam.frame_size
# scf = torch.min(inp_size / torch.FloatTensor([frame_size]), 1)[0]
outvid = False
if args.save_out != '':
outvid = True
codec = cv2.VideoWriter_fourcc(*'mp4v')
print((inp_dets * 2, inp_dets * 2))
writer = cv2.VideoWriter(args.save_out, codec, 25, (inp_dets * 2, inp_dets * 2))
fps_time = 0
f = 0
while cam.grabbed():
f += 1
frame = cam.getitem()
image = frame.copy()
# Detect humans bbox in the frame with detector model.
detected = detect_model.detect(frame, need_resize=False, expand_bb=10)
# Predict each tracks bbox of current frame from previous frames information with Kalman filter.
tracker.predict()
# Merge two source of predicted bbox together.
for track in tracker.tracks:
det = torch.tensor([track.to_tlbr().tolist() + [0.5, 1.0, 0.0]], dtype=torch.float32)
detected = torch.cat([detected, det], dim=0) if detected is not None else det
detections = [] # List of Detections object for tracking.
if detected is not None:
# detected = non_max_suppression(detected[None, :], 0.45, 0.2)[0]
# Predict skeleton pose of each bboxs.
poses = pose_model.predict(frame, detected[:, 0:4], detected[:, 4])
# Create Detections object.
detections = [Detection(kpt2bbox(ps['keypoints'].numpy()),
np.concatenate((ps['keypoints'].numpy(),
ps['kp_score'].numpy()), axis=1),
ps['kp_score'].mean().numpy()) for ps in poses]
# VISUALIZE.
if args.show_detected:
for bb in detected[:, 0:5]:
frame = cv2.rectangle(frame, (bb[0], bb[1]), (bb[2], bb[3]), (0, 0, 255), 1)
# Update tracks by matching each track information of current and previous frame or
# create a new track if no matched.
tracker.update(detections)
# Predict Actions of each track.
for i, track in enumerate(tracker.tracks):
if not track.is_confirmed():
continue
track_id = track.track_id
bbox = track.to_tlbr().astype(int)
center = track.get_center().astype(int)
action = 'pending..'
clr = (0, 255, 0)
# Use 30 frames time-steps to prediction.
if len(track.keypoints_list) == 30:
pts = np.array(track.keypoints_list, dtype=np.float32)
out = action_model.predict(pts, frame.shape[:2])
action_name = action_model.class_names[out[0].argmax()]
action = '{}: {:.2f}%'.format(action_name, out[0].max() * 100)
if action_name == 'Fall Down':
clr = (255, 0, 0)
elif action_name == 'Lying Down':
clr = (255, 200, 0)
# VISUALIZE.
if track.time_since_update == 0:
if args.show_skeleton:
frame = draw_single(frame, track.keypoints_list[-1])
frame = cv2.rectangle(frame, (bbox[0], bbox[1]), (bbox[2], bbox[3]), (0, 255, 0), 1)
frame = cv2.putText(frame, str(track_id), (center[0], center[1]), cv2.FONT_HERSHEY_COMPLEX, 0.4, (255, 0, 0), 2)
frame = cv2.putText(frame, action, (bbox[0] + 5, bbox[1] + 15), cv2.FONT_HERSHEY_COMPLEX, 0.4, clr, 1)
# Show Frame.
frame = cv2.resize(frame, (0, 0), fx=2., fy=2.)
frame = cv2.putText(frame, '%d, FPS: %f' % (f, 1.0 / (time.time() - fps_time)), (10, 20), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 1)
frame = frame[:, :, ::-1]
fps_time = time.time()
if outvid:
writer.write(frame)
cv2.imshow('frame', frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
# Clear resource.
cam.stop()
if outvid:
writer.release()
cv2.destroyAllWindows()
參考
[1].https://arxiv.org/abs/1801.07455
[2].https://blog.csdn.net/haha0825/article/details/107192773/
[3].https://github.com/yysijie/st-gcn
原文獲取方式，掃描下方二維碼
回覆【行人摔倒檢測與跟蹤】便可獲取論文與源碼
聲明：轉載請說明出處
掃描下方二維碼關注【AI人工智能初學者】公衆號，獲取更多實踐項目源碼和論文解讀，很是期待你個人相遇，讓咱們以夢爲馬，砥礪前行！！！
點「在看」給我一朵小黃花唄
本文分享自微信公衆號 - AI人工智能初學者（ChaucerG）。
若有侵權，請聯繫 support@oschina.cn 刪除。
本文參與「OSC源創計劃」，歡迎正在閱讀的你也加入，一塊兒分享。