A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Vision based occupant detection system by monocular 3D surface reconstruction
This work presents a monocular camera based object classification system for vehicle airbag deployment control in wide and frequent illumination variations. Monochrome image sequences flashed under different illumination conditions are stabilized by the double-flash technique. Furthermore, the shadowflash method minimized cast shadows in the sequences by introducing a novel sliding n-tuple strategy. By employing the active contour model, two-dimensional information of the object is extracted based on a priori knowledge of the passenger behavior. A triplet of images, of which each image is illuminated from a different direction, are sequentially used by the photometric stereo method to recover the three-dimensional shape of the object. Utilizing both the two and three-dimensional properties of the object, a 29-dimensional feature vector is defined for the training of a neural network designed to solve a three-class problem, with the classes being forward-facing child seat, rearfacing child seat, and adult. The system is tested on a database of over 84,000 frames collected from a wide range of objects in various illumination conditions. A classification accuracy of 98.9% was achieved within the decision-time limit of three seconds.
Vision based occupant detection system by monocular 3D surface reconstruction
This work presents a monocular camera based object classification system for vehicle airbag deployment control in wide and frequent illumination variations. Monochrome image sequences flashed under different illumination conditions are stabilized by the double-flash technique. Furthermore, the shadowflash method minimized cast shadows in the sequences by introducing a novel sliding n-tuple strategy. By employing the active contour model, two-dimensional information of the object is extracted based on a priori knowledge of the passenger behavior. A triplet of images, of which each image is illuminated from a different direction, are sequentially used by the photometric stereo method to recover the three-dimensional shape of the object. Utilizing both the two and three-dimensional properties of the object, a 29-dimensional feature vector is defined for the training of a neural network designed to solve a three-class problem, with the classes being forward-facing child seat, rearfacing child seat, and adult. The system is tested on a database of over 84,000 frames collected from a wide range of objects in various illumination conditions. A classification accuracy of 98.9% was achieved within the decision-time limit of three seconds.
Vision based occupant detection system by monocular 3D surface reconstruction
Yoon, J.J. (author) / Koch, C. (author) / Ellis, T.J. (author)
2004-01-01
559267 byte
Conference paper
Electronic Resource
English
Real-time 3D features reconstruction through monocular vision
| Springer Verlag | 2010
High Performance Sensor Fusion Architecture for Vision-Based Occupant Detection
| British Library Conference Proceedings | 2003