Object classification and detection is gradually becoming the major element of image analysis and computer vision. We conduct research to verify and improve the performance of algorithms in image classification and object detection that is aligned with the latest technology development and trend.
We use the convolutional neural network (CNN) to automatically categorize different types of images.
Identify users through the facial photo or video, and generates identification and tracking algorithms.
CCTV footages are used to distinguish human postures and behaviors and detect emergency / dangerous situations.
Quickly detect and distinguish objects from images collected in real time, such as driving a vehicle.
Deep learning-based facial and body part recognition algorithm enables intelligent identification of the same person on real time video and provides the assessment of the current situation through physical information (shoulders, arms, legs, etc).
It can detect abnormal behavior through various camera sensors and CCTV. It can be used in various scenarios such as monitoring workplace safety and emergency situation, checking blacklisted persons in immigration screening, and preventing terrorism.
The system captures, tracks and distinguishes characteristics of vehicles, roads, traffic lights, buildings, pedestrians, etc. from the real-time driving images to determine the probability of being a specific object. It ensures fast and high recognition rate for the images that are being detected while driving. The system can be applied to various solutions such as traffic detection, vehicle tracking, illegal parking recognition, and so forth by using video collection equipments such as CCTV and black box.
The system uses drone to take videos of the ground view and applies AI algorithms for the object detection. Deep learning-based object detection algorithm enables the automation of various tasks such as harbor facility monitoring, facility inspection, missing person search, marine vessel monitoring, and can also assist in the development of SAR drone and collection of urban environmental information.