SDE045  |  AMPS – Asset Management and Production System

To cater to the needs of growing data scales and cross border collaboration, AMPS is designed to streamline the process of resource management and build a more efficient pipeline for data management and content development. With AMPS, project management and collaboration is efficient and effective even at different physical locations.

Exhibitor : NTUitive Pte Ltd

Enhancement of Color Videos for Surveillance Systems

Video monitoring and recording are widely used in many applications such as surveillance camera systems and UAVs. Without processing, the video captured by the cameras are usually dull looking with some loss of details as compared to what is seen by the human eyes. Our video enhancement solution is designed to improve the video quality using various video processing techniques. The enhanced video using our solution will look sharper, with more depth and livelier colors, as well as good retention of details. Our solution is able to bring out hidden details captured by camera in night conditions or shadows.

Exhibitor : Panasonic R&D Center Singapore

SDE054  |  Free Space Technology and Its Applications to Automotive and Robotics

Vehicle safety has become paramount now, more than any of the previous five decades of automotive transport, owing to the number of vehicles plying urban roads. Hence, vehicles need to understand their surroundings in order to assist the driver in navigating the roads safely. Mobile robots, on the other hand, are equipped with the capability to autonomously transport and distribute items to a predetermined destination. There is a need to estimate the free space present in front of vehicle (robot) in order to navigate efficiently. We present a novel method to achieve robust free space estimation. 

Exhibitor : Panasonic R&D Center Singapore

SDE061  |  Visual-based Localization and Motion Flow Analysis for Autonomous Vehicles

Self-localization is a key perception technology required in Autonomous Vehicle (AV) to localize vehicle itself in the global map. In conventional methods, they rely on expensive 3D Lidar (Light detection and ranging) to achieve this task. In contrast, we propose a cost effective solution to self-localization by using only normal 2D cameras. We achieve localization accuracy of 20cm – 50cm which is very close to the use of 3D sensors. In addition, we also propose a motion flow analysis of the moving objects in driving environment in order to identify potential risk of specific moving object and enhance safety.

Exhibitor : Panasonic R&D Center Singapore