Hi all,

I have developed new S88 alternative solution which is "CameraS88".
Please watch the following video.

Camera S88 Video

This solution can use with Desktop Station. I will release this solution for free in my web page.
If you have Desktop Station(Railuino or DCC/MM2 Shield) and web cam, please try.



Best regards,
Yaasan

Hi bo,

Thank you for reply.
CameraS88 application can detect the difference between pre-registered and realtime image.
Because you can register many Monitoring areas, several locs may be recognized at several monitoring areas.

Best regards,
Yaasan

Hi all,

You can find Camera S88 app and corresponding Desktop Station 0.90 in my web page.
If you play, please run CameraS88 and also Desktop Station. You don't have to configure something of the CameraS88 in the Desktop Station. Desktop Station can automatically receive S88 sensor data from CameraS88 app. Of course, you need the registration of virtual S88 sensor in the CameraS88 app.

I'd like to show you CameraS88 processing. CameraS88 uses template matching which is a image recognizance method.

Application screen:


Dictionary Image(Template):


Online Image(Target)


Difference image:


CameraS88 divides the image by 16x16 pixels block. You can register areas based on the block(16x16).
If the area includes some different block, CameraS88 regards the area as having a locomotive.

Of course, this algorithm is very common but not the best approach. I will add the other algorithm.

Best regards,
Yaasan

Hi Francisco,

You will find the following instruction.

0. Connect Arduino with CAN-BUS Shield or DCC/MM2 Shield and web cam. If you connect 2 web cams, please check the selected camera later.
1. Run CameraS88 app. *
2. Run Desktop Station 0.9 or later *
3. Open CameraS88 window and click Config button. Please check the selected camera. If ok, close the config window.
4. Click Start button. You need to remove all of locomotives on the rail temporary. After that, move and adjust position of web cam. If finished, click Register button. The dictionary image (template) is registered. You can check the dictionary image on the config window.
5. You need to add monitoring areas like as this video.
6. If added all of monitoring areas, please check whether recognition works correctly or not.
7. Desktop Station can detect data of CameraS88 same as normal S88 decoder. Not need any configuration.

*Step 1 and 2 can change the order.

I captured the above procedure as video. See http://youtu.be/HgRJYg2c95A.

Instrunction of mfx recognization describes at How to add mfx locomotives. However, mfx locomotives are many habits. I can't check all behavior of mfx locomotives. Please note.

Best regards,
Yaasan

Edited by user 30 April 2014 03:45:02(UTC)  | Reason: Not specified

Hi franciscohg,

I improve the delay problem. Please download the latest version.

http://yaasan.mints.ne.jp/bin/CameraS88_R2.ZIP

You can adjust skip level which large value means skip rates.

Best regards,
Yaasan

Hi Christoffer,

Do you have a marklin MS2? If you have MS2, you use Railuino kits.

Components:
marklin MS2, 60113 (Track box), AC adapter (66361), locomotives, CTrack rails.
Railuino cable kit : http://www.watterott.com...ge=product&info=2872
CANdiy Shield : http://www.watterott.com/en/Arduino-CANdiy-Shield
Arduino UNO R3: http://www.watterott.com/en/Arduino-Uno
USB cable, Windows PC, of course web cams too.

Software (you can get all of the following softwares for free):
Arduino IDE : http://arduino.cc/en/main/software
Desktop Station, Camera S88: http://yaasan.mints.ne.jp/download.html

The detail describes at http://yaasan.mints.ne.jp/ready.html.

If you have questions, please tell me not to hesitate.

Best regards,
Yaasan