Estimating dimensions of free-swimming fish using 3-D point distribution models
Computer Vision and Image Understanding - Special issue on underwater computer vision and pattern recognition
IEEE Transactions on Pattern Analysis and Machine Intelligence
Alive Fishes Species Characterization from Video Sequences
Proceedings of the Joint IAPR International Workshop on Structural, Syntactic, and Statistical Pattern Recognition
The effect of texture representations on AAM performance
ICPR '06 Proceedings of the 18th International Conference on Pattern Recognition - Volume 02
A background robust active appearance model using active contour technique
Pattern Recognition
Fast Simplex Optimization for Active Appearance Model
PSIVT '09 Proceedings of the 3rd Pacific Rim Symposium on Advances in Image and Video Technology
Shape and Texture Based Classification of Fish Species
SCIA '09 Proceedings of the 16th Scandinavian Conference on Image Analysis
Automatic fish classification for underwater species behavior understanding
Proceedings of the first ACM international workshop on Analysis and retrieval of tracked events and motion in imagery streams
Illumination invariant face alignment using multi-band active appearance model
PReMI'05 Proceedings of the First international conference on Pattern Recognition and Machine Intelligence
Hi-index | 0.00 |
This paper presents an original Augmented-Reality system to automatically identify aquarium fish species, providing a rich multimedia experience to customers. Our goal is to replace the signs placed near tanks in aquariums with a smartphone application based on image-processing. Our system is grounded on the Active Appearance Model for fish texture sampling. This paper also introduces a novel AAM matching function that measures the superimposition degree of the AAM instance edges and the targets' edges. The newly defined function significantly improves the AAM matching performance on textureless targets without modifying the computational cost. We evaluate our identification algorithm quantitatively on a comprehensive synthetic data set of static images, whereas we evaluate the usability of our AR system in real conditions qualitatively. It yields a 94% correct-identification rate on 15 species and runs up to 15 frames per second on an iPod Touch 4G, ensuring a satisfying user experience.