Neural recognition - EU plates

[erkobg]

Hello, First - thank you a lot for sharing such a great project with us. Is there a way to train the neural network to recognize different EU plates? Particularly the Bulgarian ones? They thus take the form X(X) NNNN YY, where X is the province code, and Y is the series. Examples: https://upload.wikimedia.org/wikipedia/commons/e/e9/Bulgaria-automobile-license-plate_for_eu.png

https://upload.wikimedia.org/wikipedia/commons/thumb/5/5e/Bulgarian_license_plate.JPG/320px-Bulgarian_license_plate.JPG

Thanks and regards!

[anhtuan98]

Hi Erkobg, Thanks for your interest. We are training the general plates that can apply to any country has rectangle plates. However, it is a variations for different country. It would be nice if you can contribute sample plates (2000+ images) so we can train our system to recognise it.

Best wishes

Tuan Nguyen, Ph.D. Dean of The Faculty of Computer network & Communication University of Information Technology Km20 Hanoi Highway, Linh Trung, Thu Duc, HCMC, Vietnam Cell:+84 932215030 Homepage: http://www.uit.edu.vn/~tuanna

On Fri, Dec 11, 2015 at 7:47 PM, erkobg notifications@github.com wrote:

Hello, First - thank you a lot for sharing such a great project with us Is there a way to train the neural network to recognize different EU plates? Particularly the Bulgarian ones? They thus take the form X(X) NNNN YY, where X is the province code, and Y is the series Examples:

https://uploadwikimediaorg/wikipedia/commons/e/e9/Bulgaria-automobile-license-plate_for_eupng

https://uploadwikimediaorg/wikipedia/commons/thumb/5/5e/Bulgarian_license_plateJPG/320px-Bulgarian_license_plateJPG

Thanks and regards!

— Reply to this email directly or view it on GitHub https://github.com/anhtuan98/uitanpr2/issues/4.

[erkobg]

Hello, thanks for the replay, I am currently generating images from different traffic recordings I have. Can you share email so I can send you them? I am trying to collect plates in different angles, light etc. Please let me know if you have specific expectations.

Regards!

[linhadiretalipe]

Hi, thanks a lot for share these codes. How can I do to training it to read plates from Brazil? We have 2 kinds of plate. One kind for cars and others for motorcicles. For cars is AAA NNNN in horizontal. For motorcicles is AAA under it NNNN as you can see in the links below.

Cars plate http://www.carrosnaweb.com.br/imgsite%5Cplaca-carro.jpg

Motorcicles plate http://www.placnord.com.br/placas/p18_g.jpg

regards,

[fernando-antonio]

@linhadiretalipe consegui progredir no projeto com as placas do Brasil?

[henriqueamaral]

@fernandoc3po , @linhadiretalipe , caso tenham sucesso nas placas do Brasil, me informem por favor.

[fernando-antonio]

@henriqueamaral tenho uma base de placas do Brasil, o @linhadiretalipe desistiu do projeto. Ainda não sei como realizar o treinamento das redes neurais.

[henriqueamaral]

hum, ok. vou estudar o treinamento e vamos nos falando.

Em 16 de junho de 2016 11:14, Fernando Antonio notifications@github.com escreveu:

@henriqueamaral https://github.com/henriqueamaral tenho uma base de placas do Brasil, o @linhadiretalipe https://github.com/linhadiretalipe desistiu do projeto. Ainda não sei como realizar o treinamento das redes neurais.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/anhtuan98/uitanpr2/issues/4#issuecomment-226497610, or mute the thread https://github.com/notifications/unsubscribe/ATBy8DdkpCseV6V5Ghudft_5ZksGSSxxks5qMVo1gaJpZM4Gzi_2 .