Driver Monitoring System(DMS)
With Deep Learning Concept
I. Product description
DMS system:
※ DMS is applied to monitor vehicle’s internal status, including driver’s behavior and status, to execute risks prediction and precaution, thus to minimize the accident happening and make driving more safer.
※ With continuous functions upgrade and optimization, diversified developing of interaction modes and multi-mode integration of in-car perception, DMS becomes more and more important and carries more functions to achieve intelligent car through embedded artificial intelligence.
II. System Components
III. System Specs
CMOS Sensor |
Mono IR |
Resolution |
1280 x 720 |
HFOV |
60~70° |
IR LED |
940nm |
Dynamic Range |
64dB |
SNR |
38dB |
Optical Distortion |
<17% |
Working Voltage |
9V~16V |
Working Current |
<500mA |
Working Temperature |
-40℃~85℃ |
Storage Temperature |
-40℃~95℃ |
IV. System Principle
Based on traditional computer vision and learning model, facial and cognitive state recognition technology consists of 3 steps:
Monocular IR camera → face recognition → use CNN model to position and track the eyes/mouth on the face → voice warning
Software workflow
V. System Functions
Function Voice |
Warning Sentence |
Accuracy |
Fatigue Warning |
Beware of fatigue! |
95% |
Distraction Warning |
Please drive carefully! |
96% |
Smoking Warning |
No smoking! |
98% |
Phone Call Warning |
No calls! |
98% |
Driver Abnormal Warning |
Face lost! |
96% |
VI. Installation
Below is the installation position recommended.