# 窮人也玩得起的local物件辨識
###### tags: `FRC` `Object Detection` `Raspberry Pi`
## Step 1: Set up hardware(Pi 3)
1) Wires: HDMI, Micro USB, RJ45, USB
2) Camera: Pi camera is the Default Option, other cameras can be used also
3) Micro SD card: Write code after plugging it into the SD card port on the computer
4) Coral USB Accelerator: Increases the inference speed by ~10x

## Step 2: Install the TensorFlow Lite runtime
### TensorFlow Lite converter
>TensorFlow model => TensorFlow Lite FlatBuffer file (.tflite)
Supports SavedModel directories, tf.keras models, and concrete functions
### Pip
> 讓安裝Python套件更簡單

### 轉換方式

converter = lite.TFLiteConverter.from_keras_model(model)
tflite_model = converter.convert()

## Reference

### [tf.lite.Interpreter](https://www.tensorflow.org/lite/guide/python)
### [Tutorial](https://github.com/tensorflow/examples/blob/master/lite/examples/object_detection/raspberry_pi/README.md)