34 lines
2.2 KiB
Markdown
34 lines
2.2 KiB
Markdown
## 如何运行
|
||
|
||
0. 环境与依赖：
|
||
- python 3.7
|
||
- tensorflow 2.0.0b0
|
||
- pillow(PIL) 4.3.0
|
||
1. 下载数据集：[The EMNIST Dataset](https://www.nist.gov/itl/products-and-services/emnist-dataset)
|
||
2. 解压数据集，选取其中的
|
||
- emnist-letters-train-labels-idx1-ubyte.gz
|
||
- emnist-letters-train-images-idx3-ubyte.gz
|
||
- emnist-letters-test-labels-idx1-ubyte.gz
|
||
- emnist-letters-test-images-idx3-ubyte.gz
|
||
复制到 `data_set_emnist_letters` 目录中（已放入）
|
||
3. 运行 `src/letters/letter_train.py` 开始模型训练
|
||
4. 运行 `src/letters/letter_predict.py` ，对 `english_images` 下的所有图片进行识别，并逐一显示结果
|
||
|
||
## 附 字符编码表（UTF-8）
|
||
|
||
| **字符** | **编码10进制** | **编码16进制** | **Unicode编码10进制** | **Unicode编码16进制** |
|
||
| -------- | -------------- | -------------- | --------------------- | --------------------- |
|
||
| a | 97 | 61 | 97 | 61 |
|
||
| b | 98 | 62 | 98 | 62 |
|
||
| c | 99 | 63 | 99 | 63 |
|
||
| d | 100 | 64 | 100 | 64 |
|
||
| e | 101 | 65 | 101 | 65 |
|
||
| f | 102 | 66 | 102 | 66 |
|
||
| z | 122 | 7A | 122 | 7A |
|
||
| A | 65 | 41 | 65 | 41 |
|
||
| B | 66 | 42 | 66 | 42 |
|
||
| C | 67 | 43 | 67 | 43 |
|
||
| D | 68 | 44 | 68 | 44 |
|
||
| E | 69 | 45 | 69 | 45 |
|
||
| F | 70 | 46 | 70 | 46 |
|
||
| Z | 90 | 5A | 90 | 5A | |