Two-bit weights are enough to solve vehicle license number recognition problem

Lisa, Ferran; Carrabina, Jordi; Pérez-Vicente, Conrad J.; Avellana, Narcís; Valderrama, Elena

doi:10.1109/icnn.1993.298735

Cited by 5 publications

(2 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The recognition method used is similar to the methods for recognizing the characters on license plates, including neural-network methods (Lisa et al 1993, Kertesz et al 1994, Tindall 1995, Pandya and Macy 1996, Raus and Kreft 1996, Parisi et al 1998, template-matching methods (Otsu 1979, Comelli et al 1995, Hegt et al 1998, and structural methods (Pan et al 2004, Lin et al 2010. Most of the neural-network methods designed for recognizing the characters on license plates implement multilayer perception in the network structure to coordinate different types of learning methods (Lisa et al 1993). The popular neuralnetwork method offers adjustability, nonlinear processing, parallel calculation, and antistatic ability (Pandya and Macy 1996).…”

Section: Introductionmentioning

confidence: 99%

Interactive image tools for autonomous spelling robot

Lin

Hsueh

et al. 2012

Journal of the Chinese Institute of Engineers

View full text Add to dashboard Cite

This research proposes a simple but reliable command card encoding system, image recognition, and interpretation techniques for reading the command cards and a letter cube image recognition system for a twoarmed autonomous spelling and cube-stacking robot. The command cards are used for issuing commands to the robot so that people can ask the robot to execute specific actions. By using unique identification marks and encoding systems, the command cards can be quickly recognized and correctly interpreted by the robot. The card is still readable even when one corner is covered by fingers. The letter cube recognition system developed in this research uses a white background inside a black square frame as the target feature. This system can identify a letter cube viewed from different angles and distances, and then recognize the letters printed on the cube by a template matching method. The two vision-recognition systems serve as reliable image interaction tools for the multifunctional robot to receive commands from the card and execute the requested actions in a fully autonomous manner.

show abstract

Section: Introductionmentioning

confidence: 99%

Interactive image tools for autonomous spelling robot

Lin

Hsueh

et al. 2012

Journal of the Chinese Institute of Engineers

View full text Add to dashboard Cite

show abstract

“…Therefore, the study is based on introducing the recognition of the characters on license plates. As for the related domestic and international documents, the most common methods are Neural Network method [1][2][3][4][5][6], Template matching methods [7][8], and Structural method. Most of the Neural Network Method that is designed in recognizing the characters on the license plate implement the MLP -Multi-Layer Perception to be its network structure to coordinate with different types of learning the mathematical calculation method.…”

Section: Introductionmentioning

confidence: 99%

Recognition technique for character cube stacking robot

Lin

Hsueh

2008

2008 10th International Conference on Control, Automation, Robotics and Vision

View full text Add to dashboard Cite

This article is about designing a robot that is able to interact and play with the alphabet blocks as well as a visual recognition technique for the robots to be able to recognize the blocks and the letters on the blocks. The background color of the blocks is white, and the letters on the blocks are in black, which creates a characteristic of isolating each target from the group. By using the quadrangle detection method mentioned in the article to obtain the position of each quadrangle face of each letter cube, and after filtering the shapes and other characteristics in order to obtain the possible area, the letter cube is able to be positioned and recognized from the background after implement template matching and sifting out the ratio scale of the color area. As for the interaction with the robot, the article has also presented the complete operation procedure from the recognition of the letter blocks' position, grabbing the blocks, locating the stocking platform, the correction of the position of the blocks, and piling the blocks. This allows the robot to have the interactive ability to search, grab, rotate, and pile the letter blocks based on the order of the inputted word.

show abstract

Car license plate recognition with neural networks and fuzzy logic

Nijhuis

Brugge

Helmholt

et al. 1995

Proceedings of ICNN'95 - International Conference on Neural Networks

140

View full text Add to dashboard Cite

A car license plate recognition system (CLPR-system) has been developed to identib vehicles by the contents of their license plate for speed-limit enforcement. This type of application puts high demands on the reliability of the CLPR-system. A combination of neural andfizzy techniques is used to guarantee a very low error rate at an acceptable recognition rate. First experiments along highways in the Netherlands show that the system has an error rate of 0.02% at a recognition rate of 98.51%. These results are also compared with otherpublished CLPR-systems.

show abstract

Two-bit weights are enough to solve vehicle license number recognition problem

Cited by 5 publications

References 2 publications

Interactive image tools for autonomous spelling robot

Interactive image tools for autonomous spelling robot

Recognition technique for character cube stacking robot

Car license plate recognition with neural networks and fuzzy logic

Contact Info

Product

Resources

About