24 octobre 2019

DEEP LEARNING COURSES AT MASTER LEVEL BY NEXYAD MEMBER OF MOVEO GROUPEMENT ADAS : TEACHING CHAIR WITH INSA ROUEN

DEEP LEARNING COURSES AT MASTER LEVEL BY NEXYAD MEMBER OF MOVEO GROUPEMENT ADAS : TEACHING CHAIR WITH INSA ROUEN NEXYAD CEO TEACHING DEEP LEARNING TO MASTER STUDENT OFNORMANDY ENGINEERING SCHOOL INSA ROUEN INSA ROUEN and MOVEO GROUPEMENT ADAS created a teaching chair on intelligent vehicles (ADAS, Autonomous Driving, connected vehicle, ...). Last week, it was all about deep learning applied to real world applications : . concepts . how to build NN architecture ? . how to build learning database and validation database ? .... [Lire la suite]

21 février 2019

FREE SPACE DETECTION BY CAMERA ON REAL WORLD AREAS NEAR PARIS WITH POOR QUALITY OF BITUMEN ON THE ROAD : CRACKS, PATCHES, ...

FREE SPACE DETECTION BY CAMERA ON REAL WORLD AREAS NEAR PARIS WITH POOR QUALITY OF BITUMEN ON THE ROAD : CRACKS, PATCHES, ...   Some companies try to detect freespace with a regular camera, without IR (SWIR etc …), without fusion with LIDAR, just because it allows low cost applications especially for aftermarket. Here is an example of the NEXYAD module RoadNex. One can see that there are a lot of cracks, patches (highlights and dark zones), and that RoadNex still detects free space in front of the car. Especially useful in... [Lire la suite]
Posté par NEXYAD à 00:00 - - Commentaires [0] - Permalien [#]
Tags : , , , , , , , , , ,
25 octobre 2018

Free Space detection in real time on standard architectures such as smartphones

Free Space detection in real time on standard architectures such as smartphones Here is the demo of RoadNex free space detection on a regular smartphone. Free space is detected in real time. On this version, "distant  sidewalks" are integrated into free space (it is an option) because sidewalks without pedestrians may be a option to avoid accident if it is not possible to brake as short as need and is the street is full of users. No need for a NVIDIA heavy computing architecture, no need for a MOBILEYE chip, etc ... Only the... [Lire la suite]
23 septembre 2018

NEXYAD in SHANGHAI FOR AUTONOMOUS DRIVING CONGRESS

NEXYAD in SHANGHAI FOR AUTONOMOUS DRIVING CONGRESS Nexyad on wall of fame at 2018 World Autonomous Vehicle Conference in Shanghai. Dont' miss Groupement ADAS corner september 18-19 ! NEXYAD shows : Real time driving risk computing, road detection, obstacle detection, visibility measurement #autonomousvehicles #conferences #avs
23 février 2018

Low CPU consumption road detection and free space detection running in real time on a regular smartphone : RoadNex

Low CPU consumption road detection and free space detection running in real time on a regular smartphone : RoadNex The company NEXYAD showed to its partners and customers at CES 2018 in Las Vegas the RoadNex road detection system running on a regular smartphone (using CPU and camera of the smartphone). RoadNex detectes two things, independently : . borders of the lane : borders may be road markings, other cars, construction area signs, pedestrians, bicycle, ... . free space : part of the world ahead that is flat and drivable... [Lire la suite]
02 février 2018

Free space detection with RoadNex : real time implementation on a regular smartphone

Free space detection with RoadNex : real time implementation on a regular smartphone  Panoramic movement (the most complex for any detection algorithm) with smartphone in hand (so calibration must be very robust as heigth and angle of camera is changing all the time). The goal of this free space detection is to detect as more drivable surface as possible (including sidewalks if not too high, including security area, etc ...). Indeed, the "lane" borders are detected by a borders detection algorithm (different from free space). ... [Lire la suite]

27 octobre 2017

ADAS : bitumen free space détection with the NEXYAD road detection module RoadNex

ADAS : bitumen free space détection with the NEXYAD road detection module RoadNex And now in real time on a smartphone : https://www.youtube.com/watch?v=f5RRl_0Bz2g  Read more : http://www.nexyad.net/Automotive-Transportation  
17 octobre 2017

ADAS glossary and how the 4 innovative modules of NEXYAD can simplify everything

ADAS glossary and how the 4 innovative modules of NEXYAD can simplify everything NEXYAD has been developing 4 software modules (RoadNex, ObstaNex, VisiNex, SafetyNex) for ADAS and road safety applications. Those 4 modules are designed to be deployed in series : . smartphones . aftermarket ADAS . telematics . intelligent cars Here is the glossary of main ADAS and how thos 4 modules can be used in every case. Intersection Assistant SafetyNex estimates driving risk, onboard and in real time. SafetyNex estimates adequation... [Lire la suite]
05 septembre 2017

Use Case of computer vision ADAS system (RoadNex free space detection) : automatic predictive braking

Use Case of computer vision ADAS system (RoadNex free space detection) : automatic predictive braking NEXYAD made some tries of using their module RoadNex for predictive braking. RoadNex is a computer vision based module for road detection : . detection of drivable area borders and detection of free space (drivable area surface) :   The idea is that RoadNex is NOT and obstacle detector, but ifever the drivable area is going thinner or smaller, thenthere is a chance that an obstacle should be detected. It means that RoadNex... [Lire la suite]
14 août 2017

Free space detection using the NEXYAD road detection module RoadNex (ADAS, driverless)

Free space detection using the NEXYAD road detection module RoadNex (ADAS, driverless) NEXYAD proposes a sofwtare module called RoadNex that detects : . borders of the lane (with or without road signs : white lines etc, RoadNex works even on a desert track) . free space (area in front of the car that can be considered as "drivable") Even without any obstacle detection, it is possible then to know that there is "something" on the road or not. Note : RoadNex won't make a difference between a drawing on the road and an obstacle, but... [Lire la suite]