[OpenR8 solution] HttpServer_DeepLearning_Python_PCB (Python TensorFlow Keras Deep Learning Server for PCB Inspection)
  1. HttpServer_DeepLearning_Python_PCB


The purpose of this solution : The user selects the image through the Web page and obtains the frame that the result message of the object detects.


First, select the image on the Web page, transfer the image, server receive the picture and then use the trained model to judge, the detected object result message is sent back to the Web page. The main process used is shown in Fig. 1, Fig. 2 below.


The model detection object for this solution is the capacitor in the PCB. The schematic diagram is shown in Fig. 3 below.


Path :  File root directory/solution/HttpServer_DeepLearning_Python_PCB/model/data/.


By replacing different models, different model test results can be obtained on the Web page.


If you want to detect animals, vehicles, and so on, simply replace the model with a trained model (additional generation required).


For example, the file root directory /solution/HttpServer_DeepLearning_Python_PCB solution is to detect 20 categories, details of which can be referred to its documentation.


Fig. 1. Flow chart..png

Fig. 1. Flow chart.


Fig. 2. Flow diagram..png

Fig. 2. Flow diagram.


Fig. 3. Categories of detected objects Capacitors in the PCB..png

Fig. 3. Categories of detected objects Capacitors in the PCB.


※This solution runs in the Keras v2.1.5 version.


This example requires Python to be installed.

  1. How to use it


There are only three steps to using this solution. As follows:


  1. Load Solution.

【Run R8.exe】=>【File】=>【Open】 =>【Select solutionHttpServer_DeepLearning_Python_PCB】 =>【Click twice HttpServer_DeepLearning_Python_PCB.flow】. See Fig. 4, Fig. 5, Fig. 6, Fig. 7 below.


The loaded flow screen is shown in Fig. 8 below.


Fig. 4. Run R8 ..png

Fig. 4. 【Run R8.exe】.


Fig. 5. File Open ..png

Fig. 5. 【File】=>【Open】.


Fig. 6. Select solution ..png

Fig. 6. 【Select solutionHttpServer_DeepLearning_Python_PCB】.


Fig. 7. Click twice HttpServer_DeepLearning_Python_PCB ..png

Fig. 7. 【Click twice HttpServer_DeepLearning_Python_PCB】.


Fig. 8. Load HttpServer_DeepLearning_Python_PCB..jpg

Fig. 8. Load HttpServer_DeepLearning_Python_PCB.


  1. Run the solution.


Fig. 9. Execute the solution..png

Fig. 9. Execute the solution.


  1. Select the image you want to test through the Web page. Press submit to obtain the results of the image detection object. Figure 10 Figure 11 below.

The test picture is placed in: root directory/solution/solution/HttpServer_DeepLearning_Python_PCB/data


Fig. 10. Upload images through a Web page..png

Fig. 10. Upload images through a Web page.


Fig. 11. Detect result messages..png

Fig. 11. Detect result messages.



Recommended Article

OpenR8 Community Edition - AI Software for Everyone (Free Download)