Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into dygraph_rc

release/2.0-rc1-0
WenmuZhou 4 years ago
commit 7b53596ce6

@ -4,16 +4,18 @@
PaddleOCR旨在打造一套丰富、领先、且实用的OCR工具库助力使用者训练出更好的模型并应用落地。
**近期更新**
- 2020.12.07 [FAQ](./doc/doc_ch/FAQ.md)新增5个高频问题总数124个并且计划以后每周一都会更新欢迎大家持续关注。
- 2020.11.25 更新半自动标注工具[PPOCRLabel](./PPOCRLabel/README.md)辅助开发者高效完成标注任务输出格式与PP-OCR训练任务完美衔接。
- 2020.9.22 更新PP-OCR技术文章https://arxiv.org/abs/2009.09941
- 2020.9.19 更新超轻量压缩ppocr_mobile_slim系列模型整体模型3.5M(详见[PP-OCR Pipline](#PP-OCR)),适合在移动端部署使用。[模型下载](#模型下载)
- 2020.9.19 更新超轻量压缩ppocr_mobile_slim系列模型整体模型3.5M(详见[PP-OCR Pipeline](#PP-OCR)),适合在移动端部署使用。[模型下载](#模型下载)
- 2020.9.17 更新超轻量ppocr_mobile系列和通用ppocr_server系列中英文ocr模型媲美商业效果。[模型下载](#模型下载)
- 2020.9.17 更新[英文识别模型](./doc/doc_ch/models_list.md#英文识别模型)和[多语言识别模型](doc/doc_ch/models_list.md#多语言识别模型),已支持`德语、法语、日语、韩语`,更多语种识别模型将持续更新。
- 2020.8.26 更新OCR相关的84个常见问题及解答具体参考[FAQ](./doc/doc_ch/FAQ.md)
- 2020.8.24 支持通过whl包安装使用PaddleOCR具体参考[Paddleocr Package使用说明](./doc/doc_ch/whl.md)
- 2020.8.21 更新8月18日B站直播课回放和PPT课节2易学易用的OCR工具大礼包[获取地址](https://aistudio.baidu.com/aistudio/education/group/info/1519)
- [More](./doc/doc_ch/update.md)
## 特性
- PPOCR系列高质量预训练模型准确的识别效果
@ -48,15 +50,14 @@ PaddleOCR旨在打造一套丰富、领先、且实用的OCR工具库助力
- 代码体验:从[快速安装](./doc/doc_ch/installation.md) 开始
<a name="模型下载"></a>
## PP-OCR 1.1系列模型列表9月17日更新
## PP-OCR 2.0系列模型列表(更新中
| 模型简介 | 模型名称 |推荐场景 | 检测模型 | 方向分类器 | 识别模型 |
| ------------ | --------------- | ----------------|---- | ---------- | -------- |
| 中英文超轻量OCR模型8.1M | ch_ppocr_mobile_v1.1_xx |移动端&服务器端|[推理模型](https://paddleocr.bj.bcebos.com/20-09-22/mobile/det/ch_ppocr_mobile_v1.1_det_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/20-09-22/mobile/det/ch_ppocr_mobile_v1.1_det_train.tar)|[推理模型](https://paddleocr.bj.bcebos.com/20-09-22/cls/ch_ppocr_mobile_v1.1_cls_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/20-09-22/cls/ch_ppocr_mobile_v1.1_cls_train.tar) |[推理模型](https://paddleocr.bj.bcebos.com/20-09-22/mobile/rec/ch_ppocr_mobile_v1.1_rec_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/20-09-22/mobile/rec/ch_ppocr_mobile_v1.1_rec_pre.tar) |
| 中英文通用OCR模型155.1M |ch_ppocr_server_v1.1_xx|服务器端 |[推理模型](https://paddleocr.bj.bcebos.com/20-09-22/server/det/ch_ppocr_server_v1.1_det_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/20-09-22/server/det/ch_ppocr_server_v1.1_det_train.tar) |[推理模型](https://paddleocr.bj.bcebos.com/20-09-22/cls/ch_ppocr_mobile_v1.1_cls_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/20-09-22/cls/ch_ppocr_mobile_v1.1_cls_train.tar) |[推理模型](https://paddleocr.bj.bcebos.com/20-09-22/server/rec/ch_ppocr_server_v1.1_rec_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/20-09-22/server/rec/ch_ppocr_server_v1.1_rec_pre.tar) |
| 中英文超轻量压缩OCR模型3.5M | ch_ppocr_mobile_slim_v1.1_xx| 移动端 |[推理模型](https://paddleocr.bj.bcebos.com/20-09-22/mobile-slim/det/ch_ppocr_mobile_v1.1_det_prune_infer.tar) / [slim模型](https://paddleocr.bj.bcebos.com/20-09-22/mobile/lite/ch_ppocr_mobile_v1.1_det_prune_opt.nb) |[推理模型](https://paddleocr.bj.bcebos.com/20-09-22/cls/ch_ppocr_mobile_v1.1_cls_quant_infer.tar) / [slim模型](https://paddleocr.bj.bcebos.com/20-09-22/mobile/lite/ch_ppocr_mobile_v1.1_cls_quant_opt.nb)| [推理模型](https://paddleocr.bj.bcebos.com/20-09-22/mobile-slim/rec/ch_ppocr_mobile_v1.1_rec_quant_infer.tar) / [slim模型](https://paddleocr.bj.bcebos.com/20-09-22/mobile/lite/ch_ppocr_mobile_v1.1_rec_quant_opt.nb)|
| 中英文超轻量OCR模型8.1M | ch_ppocr_mobile_v2.0_xx |移动端&服务器端|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar)|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar) |[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_pre.tar) |
| 中英文通用OCR模型143M |ch_ppocr_server_v2.0_xx|服务器端 |[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_train.tar) |[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar) |[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_pre.tar) |
更多模型下载(包括多语言),可以参考[PP-OCR v1.1 系列模型下载](./doc/doc_ch/models_list.md)
更多模型下载(包括多语言),可以参考[PP-OCR v2.0 系列模型下载](./doc/doc_ch/models_list.md)
## 文档教程
- [快速安装](./doc/doc_ch/installation.md)
@ -76,9 +77,9 @@ PaddleOCR旨在打造一套丰富、领先、且实用的OCR工具库助力
- [基于Python脚本预测引擎推理](./doc/doc_ch/inference.md)
- [基于C++预测引擎推理](./deploy/cpp_infer/readme.md)
- [服务化部署](./deploy/hubserving/readme.md)
- [端侧部署](./deploy/lite/readme.md)
- [模型量化](./deploy/slim/quantization/README.md)
- [模型裁剪](./deploy/slim/prune/README.md)
- [端侧部署](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/deploy/lite/readme.md)
- [模型量化](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/deploy/slim/quantization/README.md)
- [模型裁剪](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/deploy/slim/prune/README.md)
- [Benchmark](./doc/doc_ch/benchmark.md)
- 数据集
- [通用中英文OCR数据集](./doc/doc_ch/datasets.md)
@ -96,6 +97,9 @@ PaddleOCR旨在打造一套丰富、领先、且实用的OCR工具库助力
- [许可证书](#许可证书)
- [贡献代码](#贡献代码)
***注意动态图端侧部署仍在开发中目前仅支持动态图训练、python端预测C++预测,
如果您有需要移动端部署案例或者量化裁剪,请切换到静态图分支;***
<a name="PP-OCR"></a>
## PP-OCR Pipline
<div align="center">
@ -141,6 +145,7 @@ PP-OCR是一个实用的超轻量OCR系统。主要由DB文本检测、检测框
## 贡献代码
我们非常欢迎你为PaddleOCR贡献代码也十分感谢你的反馈。
- 非常感谢 [Khanh Tran](https://github.com/xxxpsyduck) 和 [Karl Horky](https://github.com/karlhorky) 贡献修改英文文档
- 非常感谢 [zhangxin](https://github.com/ZhangXinNan)([Blog](https://blog.csdn.net/sdlypyzq)) 贡献新的可视化方式、添加.gitgnore、处理手动设置PYTHONPATH环境变量的问题
- 非常感谢 [lyl120117](https://github.com/lyl120117) 贡献打印网络结构的代码
@ -148,3 +153,6 @@ PP-OCR是一个实用的超轻量OCR系统。主要由DB文本检测、检测框
- 非常感谢 [authorfu](https://github.com/authorfu) 贡献Android和[xiadeye](https://github.com/xiadeye) 贡献IOS的demo代码
- 非常感谢 [BeyondYourself](https://github.com/BeyondYourself) 给PaddleOCR提了很多非常棒的建议并简化了PaddleOCR的部分代码风格。
- 非常感谢 [tangmq](https://gitee.com/tangmq) 给PaddleOCR增加Docker化部署服务支持快速发布可调用的Restful API服务。
- 非常感谢 [lijinhan](https://github.com/lijinhan) 给PaddleOCR增加java SpringBoot 调用OCR Hubserving接口完成对OCR服务化部署的使用。
- 非常感谢 [Mejans](https://github.com/Mejans) 给PaddleOCR增加新语言奥克西坦语Occitan的字典和语料。
- 非常感谢 [Evezerest](https://github.com/Evezerest) [ninetailskim](https://github.com/ninetailskim) [edencfc](https://github.com/edencfc) [BeyondYourself](https://github.com/BeyondYourself) [1084667371](https://github.com/1084667371) 贡献了PPOCRLabel的完整代码。

File diff suppressed because it is too large Load Diff

@ -8,7 +8,6 @@ Global:
# evaluation is run every 5000 iterations after the 4000th iteration
eval_batch_step: [0, 1000]
# if pretrained_model is saved in static mode, load_static_weights must set to True
load_static_weights: True
cal_metric_during_train: True
pretrained_model:
checkpoints:

@ -1,130 +0,0 @@
Global:
use_gpu: true
epoch_num: 1200
log_smooth_window: 20
print_batch_step: 2
save_model_dir: ./output/det_r50_vd/
save_epoch_step: 1200
# evaluation is run every 5000 iterations after the 4000th iteration
eval_batch_step: 8
# if pretrained_model is saved in static mode, load_static_weights must set to True
load_static_weights: True
cal_metric_during_train: False
pretrained_model: ./pretrain_models/ResNet50_vd_ssld_pretrained/
checkpoints:
save_inference_dir:
use_visualdl: True
infer_img: doc/imgs_en/img_10.jpg
save_res_path: ./output/det_db/predicts_db.txt
Optimizer:
name: Adam
beta1: 0.9
beta2: 0.999
learning_rate:
lr: 0.001
regularizer:
name: 'L2'
factor: 0
Architecture:
type: det
algorithm: DB
Transform:
Backbone:
name: ResNet
layers: 50
Neck:
name: FPN
out_channels: 256
Head:
name: DBHead
k: 50
Loss:
name: DBLoss
balance_loss: true
main_loss_type: DiceLoss
alpha: 5
beta: 10
ohem_ratio: 3
PostProcess:
name: DBPostProcess
thresh: 0.3
box_thresh: 0.6
max_candidates: 1000
unclip_ratio: 1.5
Metric:
name: DetMetric
main_indicator: hmean
TRAIN:
dataset:
name: SimpleDataSet
data_dir: ./detection/
file_list:
- ./detection/train_icdar2015_label.txt # dataset1
ratio_list: [1.0]
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- DetLabelEncode: # Class handling label
- IaaAugment:
augmenter_args:
- { 'type': Fliplr, 'args': { 'p': 0.5 } }
- { 'type': Affine, 'args': { 'rotate': [ -10,10 ] } }
- { 'type': Resize,'args': { 'size': [ 0.5,3 ] } }
- EastRandomCropData:
size: [ 640,640 ]
max_tries: 50
keep_ratio: true
- MakeBorderMap:
shrink_ratio: 0.4
thresh_min: 0.3
thresh_max: 0.7
- MakeShrinkMap:
shrink_ratio: 0.4
min_text_size: 8
- NormalizeImage:
scale: 1./255.
mean: [ 0.485, 0.456, 0.406 ]
std: [ 0.229, 0.224, 0.225 ]
order: 'hwc'
- ToCHWImage:
- keepKeys:
keep_keys: ['image','threshold_map','threshold_mask','shrink_map','shrink_mask'] # dataloader will return list in this order
loader:
shuffle: True
drop_last: False
batch_size: 16
num_workers: 8
EVAL:
dataset:
name: SimpleDataSet
data_dir: ./detection/
file_list:
- ./detection/test_icdar2015_label.txt
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- DetLabelEncode: # Class handling label
- DetResizeForTest:
image_shape: [736,1280]
- NormalizeImage:
scale: 1./255.
mean: [ 0.485, 0.456, 0.406 ]
std: [ 0.229, 0.224, 0.225 ]
order: 'hwc'
- ToCHWImage:
- keepKeys:
keep_keys: ['image','shape','polys','ignore_tags']
loader:
shuffle: False
drop_last: False
batch_size: 1 # must be 1
num_workers: 8

@ -11,7 +11,7 @@ Global:
load_static_weights: True
cal_metric_during_train: False
pretrained_model: ./pretrain_models/MobileNetV3_large_x0_5_pretrained
checkpoints: #./output/det_db_0.001_DiceLoss_256_pp_config_2.0b_4gpu/best_accuracy
checkpoints:
save_inference_dir:
use_visualdl: False
infer_img: doc/imgs_en/img_10.jpg

@ -11,7 +11,7 @@ Global:
load_static_weights: True
cal_metric_during_train: False
pretrained_model: ./pretrain_models/ResNet18_vd_pretrained
checkpoints: #./output/det_db_0.001_DiceLoss_256_pp_config_2.0b_4gpu/best_accuracy
checkpoints:
save_inference_dir:
use_visualdl: False
infer_img: doc/imgs_en/img_10.jpg

@ -11,7 +11,7 @@ Global:
load_static_weights: True
cal_metric_during_train: False
pretrained_model: ./pretrain_models/MobileNetV3_large_x0_5_pretrained
checkpoints: #./output/det_db_0.001_DiceLoss_256_pp_config_2.0b_4gpu/best_accuracy
checkpoints:
save_inference_dir:
use_visualdl: False
infer_img: doc/imgs_en/img_10.jpg

@ -3,7 +3,7 @@ Global:
epoch_num: 1200
log_smooth_window: 20
print_batch_step: 10
save_model_dir: ./output/det_rc/det_r50_vd/
save_model_dir: ./output/det_r50_vd/
save_epoch_step: 1200
# evaluation is run every 5000 iterations after the 4000th iteration
eval_batch_step: [5000,4000]

@ -1,106 +0,0 @@
Global:
use_gpu: false
epoch_num: 500
log_smooth_window: 20
print_batch_step: 10
save_model_dir: ./output/rec/mv3_none_bilstm_ctc/
save_epoch_step: 500
# evaluation is run every 5000 iterations after the 4000th iteration
eval_batch_step: 127
# if pretrained_model is saved in static mode, load_static_weights must set to True
load_static_weights: True
cal_metric_during_train: True
pretrained_model:
checkpoints:
save_inference_dir:
use_visualdl: False
infer_img: doc/imgs_words/ch/word_1.jpg
# for data or label process
max_text_length: 80
character_dict_path: ppocr/utils/ppocr_keys_v1.txt
character_type: 'ch'
use_space_char: False
infer_mode: False
use_tps: False
Optimizer:
name: Adam
beta1: 0.9
beta2: 0.999
learning_rate:
lr: 0.001
regularizer:
name: 'L2'
factor: 0.00001
Architecture:
type: rec
algorithm: CRNN
Transform:
Backbone:
name: MobileNetV3
scale: 0.5
model_name: small
small_stride: [ 1, 2, 2, 2 ]
Neck:
name: SequenceEncoder
encoder_type: fc
hidden_size: 96
Head:
name: CTC
fc_decay: 0.00001
Loss:
name: CTCLoss
PostProcess:
name: CTCLabelDecode
Metric:
name: RecMetric
main_indicator: acc
TRAIN:
dataset:
name: SimpleDataSet
data_dir: ./rec
file_list:
- ./rec/train.txt # dataset1
ratio_list: [ 0.4,0.6 ]
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecAug:
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader will return list in this order
loader:
batch_size: 256
shuffle: True
drop_last: True
num_workers: 8
EVAL:
dataset:
name: SimpleDataSet
data_dir: ./rec
file_list:
- ./rec/val.txt
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader will return list in this order
loader:
shuffle: False
drop_last: False
batch_size: 256
num_workers: 8

@ -1,104 +0,0 @@
Global:
use_gpu: false
epoch_num: 500
log_smooth_window: 20
print_batch_step: 10
save_model_dir: ./output/rec/res34_none_bilstm_ctc/
save_epoch_step: 500
# evaluation is run every 5000 iterations after the 4000th iteration
eval_batch_step: 127
# if pretrained_model is saved in static mode, load_static_weights must set to True
load_static_weights: True
cal_metric_during_train: True
pretrained_model:
checkpoints:
save_inference_dir:
use_visualdl: False
infer_img: doc/imgs_words/ch/word_1.jpg
# for data or label process
max_text_length: 80
character_dict_path: ppocr/utils/ppocr_keys_v1.txt
character_type: 'ch'
use_space_char: False
infer_mode: False
use_tps: False
Optimizer:
name: Adam
beta1: 0.9
beta2: 0.999
learning_rate:
lr: 0.001
regularizer:
name: 'L2'
factor: 0.00001
Architecture:
type: rec
algorithm: CRNN
Transform:
Backbone:
name: ResNet
layers: 34
Neck:
name: SequenceEncoder
encoder_type: fc
hidden_size: 96
Head:
name: CTC
fc_decay: 0.00001
Loss:
name: CTCLoss
PostProcess:
name: CTCLabelDecode
Metric:
name: RecMetric
main_indicator: acc
TRAIN:
dataset:
name: SimpleDataSet
data_dir: ./rec
file_list:
- ./rec/train.txt # dataset1
ratio_list: [ 0.4,0.6 ]
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecAug:
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader will return list in this order
loader:
batch_size: 256
shuffle: True
drop_last: True
num_workers: 8
EVAL:
dataset:
name: SimpleDataSet
data_dir: ./rec
file_list:
- ./rec/val.txt
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader will return list in this order
loader:
shuffle: False
drop_last: False
batch_size: 256
num_workers: 8

@ -1,41 +1,38 @@
Global:
use_gpu: false
epoch_num: 500
use_gpu: true
epoch_num: 72
log_smooth_window: 20
print_batch_step: 10
save_model_dir: ./output/rec/res34_none_none_ctc/
save_epoch_step: 500
# evaluation is run every 5000 iterations after the 4000th iteration
eval_batch_step: 127
save_model_dir: ./output/rec/ic15/
save_epoch_step: 3
# evaluation is run every 2000 iterations
eval_batch_step: [0, 2000]
# if pretrained_model is saved in static mode, load_static_weights must set to True
load_static_weights: True
cal_metric_during_train: True
pretrained_model:
checkpoints:
save_inference_dir:
use_visualdl: False
infer_img: doc/imgs_words/ch/word_1.jpg
infer_img: doc/imgs_words_en/word_10.png
# for data or label process
max_text_length: 80
character_dict_path: ppocr/utils/ppocr_keys_v1.txt
character_type: 'ch'
use_space_char: False
character_dict_path: ppocr/utils/ic15_dict.txt
character_type: ch
max_text_length: 25
infer_mode: False
use_tps: False
use_space_char: False
Optimizer:
name: Adam
beta1: 0.9
beta2: 0.999
learning_rate:
lr: 0.001
lr:
learning_rate: 0.0005
regularizer:
name: 'L2'
factor: 0.00001
factor: 0
Architecture:
type: rec
model_type: rec
algorithm: CRNN
Transform:
Backbone:
@ -43,10 +40,11 @@ Architecture:
layers: 34
Neck:
name: SequenceEncoder
encoder_type: reshape
encoder_type: rnn
hidden_size: 256
Head:
name: CTC
fc_decay: 0.00001
name: CTCHead
fc_decay: 0
Loss:
name: CTCLoss
@ -58,46 +56,42 @@ Metric:
name: RecMetric
main_indicator: acc
TRAIN:
Train:
dataset:
name: SimpleDataSet
data_dir: ./rec
file_list:
- ./rec/train.txt # dataset1
ratio_list: [ 0.4,0.6 ]
data_dir: ./train_data/
label_file_list: ["./train_data/train_list.txt"]
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecAug:
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader will return list in this order
image_shape: [3, 32, 100]
- KeepKeys:
keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order
loader:
batch_size: 256
shuffle: True
batch_size_per_card: 256
drop_last: True
num_workers: 8
EVAL:
Eval:
dataset:
name: SimpleDataSet
data_dir: ./rec
file_list:
- ./rec/val.txt
data_dir: ./train_data/
label_file_list: ["./train_data/train_list.txt"]
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader will return list in this order
image_shape: [3, 32, 100]
- KeepKeys:
keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order
loader:
shuffle: False
drop_last: False
batch_size: 256
num_workers: 8
batch_size_per_card: 256
num_workers: 4

@ -1,9 +0,0 @@
*.iml
.gradle
/local.properties
/.idea/*
.DS_Store
/build
/captures
.externalNativeBuild

@ -1,19 +0,0 @@
# 如何快速测试
### 1. 安装最新版本的Android Studio
可以从https://developer.android.com/studio 下载。本Demo使用是4.0版本Android Studio编写。
### 2. 按照NDK 20 以上版本
Demo测试的时候使用的是NDK 20b版本20版本以上均可以支持编译成功。
如果您是初学者可以用以下方式安装和测试NDK编译环境。
点击 File -> New ->New Project 新建 "Native C++" project
### 3. 导入项目
点击 File->New->Import Project... 然后跟着Android Studio的引导导入
# 获得更多支持
前往[端计算模型生成平台EasyEdge](https://ai.baidu.com/easyedge/app/open_source_demo?referrerUrl=paddlelite),获得更多开发支持:
- Demo APP可使用手机扫码安装方便手机端快速体验文字识别
- SDK模型被封装为适配不同芯片硬件和操作系统SDK包括完善的接口方便进行二次开发

@ -1,94 +0,0 @@
import java.security.MessageDigest
apply plugin: 'com.android.application'
android {
compileSdkVersion 29
defaultConfig {
applicationId "com.baidu.paddle.lite.demo.ocr"
minSdkVersion 23
targetSdkVersion 29
versionCode 1
versionName "1.0"
testInstrumentationRunner "android.support.test.runner.AndroidJUnitRunner"
externalNativeBuild {
cmake {
cppFlags "-std=c++11 -frtti -fexceptions -Wno-format"
arguments '-DANDROID_PLATFORM=android-23', '-DANDROID_STL=c++_shared' ,"-DANDROID_ARM_NEON=TRUE"
}
}
ndk {
// abiFilters "arm64-v8a", "armeabi-v7a"
abiFilters "arm64-v8a", "armeabi-v7a"
ldLibs "jnigraphics"
}
}
buildTypes {
release {
minifyEnabled false
proguardFiles getDefaultProguardFile('proguard-android-optimize.txt'), 'proguard-rules.pro'
}
}
externalNativeBuild {
cmake {
path "src/main/cpp/CMakeLists.txt"
version "3.10.2"
}
}
}
dependencies {
implementation fileTree(include: ['*.jar'], dir: 'libs')
implementation 'androidx.appcompat:appcompat:1.1.0'
implementation 'androidx.constraintlayout:constraintlayout:1.1.3'
testImplementation 'junit:junit:4.12'
androidTestImplementation 'com.android.support.test:runner:1.0.2'
androidTestImplementation 'com.android.support.test.espresso:espresso-core:3.0.2'
}
def archives = [
[
'src' : 'https://paddlelite-demo.bj.bcebos.com/libs/android/paddle_lite_libs_v2_6_1.tar.gz',
'dest': 'PaddleLite'
],
[
'src' : 'https://paddlelite-demo.bj.bcebos.com/libs/android/opencv-4.2.0-android-sdk.tar.gz',
'dest': 'OpenCV'
],
[
'src' : 'https://paddleocr.bj.bcebos.com/deploy/lite/ocr_v1_for_cpu.tar.gz',
'dest' : 'src/main/assets/models/ocr_v1_for_cpu'
]
]
task downloadAndExtractArchives(type: DefaultTask) {
doFirst {
println "Downloading and extracting archives including libs and models"
}
doLast {
// Prepare cache folder for archives
String cachePath = "cache"
if (!file("${cachePath}").exists()) {
mkdir "${cachePath}"
}
archives.eachWithIndex { archive, index ->
MessageDigest messageDigest = MessageDigest.getInstance('MD5')
messageDigest.update(archive.src.bytes)
String cacheName = new BigInteger(1, messageDigest.digest()).toString(32)
// Download the target archive if not exists
boolean copyFiles = !file("${archive.dest}").exists()
if (!file("${cachePath}/${cacheName}.tar.gz").exists()) {
ant.get(src: archive.src, dest: file("${cachePath}/${cacheName}.tar.gz"))
copyFiles = true; // force to copy files from the latest archive files
}
// Extract the target archive if its dest path does not exists
if (copyFiles) {
copy {
from tarTree("${cachePath}/${cacheName}.tar.gz")
into "${archive.dest}"
}
}
}
}
}
preBuild.dependsOn downloadAndExtractArchives

@ -1,21 +0,0 @@
# Add project specific ProGuard rules here.
# You can control the set of applied configuration files using the
# proguardFiles setting in build.gradle.
#
# For more details, see
# http://developer.android.com/guide/developing/tools/proguard.html
# If your project uses WebView with JS, uncomment the following
# and specify the fully qualified class name to the JavaScript interface
# class:
#-keepclassmembers class fqcn.of.javascript.interface.for.webview {
# public *;
#}
# Uncomment this to preserve the line number information for
# debugging stack traces.
#-keepattributes SourceFile,LineNumberTable
# If you keep the line number information, uncomment this to
# hide the original source file name.
#-renamesourcefileattribute SourceFile

@ -1,26 +0,0 @@
package com.baidu.paddle.lite.demo.ocr;
import android.content.Context;
import android.support.test.InstrumentationRegistry;
import android.support.test.runner.AndroidJUnit4;
import org.junit.Test;
import org.junit.runner.RunWith;
import static org.junit.Assert.*;
/**
* Instrumented test, which will execute on an Android device.
*
* @see <a href="http://d.android.com/tools/testing">Testing documentation</a>
*/
@RunWith(AndroidJUnit4.class)
public class ExampleInstrumentedTest {
@Test
public void useAppContext() {
// Context of the app under test.
Context appContext = InstrumentationRegistry.getTargetContext();
assertEquals("com.baidu.paddle.lite.demo", appContext.getPackageName());
}
}

@ -1,39 +0,0 @@
<?xml version="1.0" encoding="utf-8"?>
<manifest xmlns:android="http://schemas.android.com/apk/res/android"
package="com.baidu.paddle.lite.demo.ocr">
<uses-permission android:name="android.permission.WRITE_EXTERNAL_STORAGE"/>
<uses-permission android:name="android.permission.READ_EXTERNAL_STORAGE"/>
<uses-permission android:name="android.permission.CAMERA"/>
<application
android:allowBackup="true"
android:icon="@mipmap/ic_launcher"
android:label="@string/app_name"
android:roundIcon="@mipmap/ic_launcher_round"
android:supportsRtl="true"
android:theme="@style/AppTheme">
<!-- to test MiniActivity, change this to com.baidu.paddle.lite.demo.ocr.MiniActivity -->
<activity android:name="com.baidu.paddle.lite.demo.ocr.MainActivity">
<intent-filter>
<action android:name="android.intent.action.MAIN"/>
<category android:name="android.intent.category.LAUNCHER"/>
</intent-filter>
</activity>
<activity
android:name="com.baidu.paddle.lite.demo.ocr.SettingsActivity"
android:label="Settings">
</activity>
<provider
android:name="androidx.core.content.FileProvider"
android:authorities="com.baidu.paddle.lite.demo.ocr.fileprovider"
android:exported="false"
android:grantUriPermissions="true">
<meta-data
android:name="android.support.FILE_PROVIDER_PATHS"
android:resource="@xml/file_paths"></meta-data>
</provider>
</application>
</manifest>

Binary file not shown.

Before

Width:  |  Height:  |  Size: 62 KiB

@ -1,117 +0,0 @@
# For more information about using CMake with Android Studio, read the
# documentation: https://d.android.com/studio/projects/add-native-code.html
# Sets the minimum version of CMake required to build the native library.
cmake_minimum_required(VERSION 3.4.1)
# Creates and names a library, sets it as either STATIC or SHARED, and provides
# the relative paths to its source code. You can define multiple libraries, and
# CMake builds them for you. Gradle automatically packages shared libraries with
# your APK.
set(PaddleLite_DIR "${CMAKE_CURRENT_SOURCE_DIR}/../../../PaddleLite")
include_directories(${PaddleLite_DIR}/cxx/include)
set(OpenCV_DIR "${CMAKE_CURRENT_SOURCE_DIR}/../../../OpenCV/sdk/native/jni")
message(STATUS "opencv dir: ${OpenCV_DIR}")
find_package(OpenCV REQUIRED)
message(STATUS "OpenCV libraries: ${OpenCV_LIBS}")
include_directories(${OpenCV_INCLUDE_DIRS})
aux_source_directory(. SOURCES)
set(CMAKE_CXX_FLAGS
"${CMAKE_CXX_FLAGS} -ffast-math -Ofast -Os"
)
set(CMAKE_CXX_FLAGS
"${CMAKE_CXX_FLAGS} -fvisibility=hidden -fvisibility-inlines-hidden -fdata-sections -ffunction-sections"
)
set(CMAKE_SHARED_LINKER_FLAGS
"${CMAKE_SHARED_LINKER_FLAGS} -Wl,--gc-sections -Wl,-z,nocopyreloc")
add_library(
# Sets the name of the library.
Native
# Sets the library as a shared library.
SHARED
# Provides a relative path to your source file(s).
${SOURCES})
find_library(
# Sets the name of the path variable.
log-lib
# Specifies the name of the NDK library that you want CMake to locate.
log)
add_library(
# Sets the name of the library.
paddle_light_api_shared
# Sets the library as a shared library.
SHARED
# Provides a relative path to your source file(s).
IMPORTED)
set_target_properties(
# Specifies the target library.
paddle_light_api_shared
# Specifies the parameter you want to define.
PROPERTIES
IMPORTED_LOCATION
${PaddleLite_DIR}/cxx/libs/${ANDROID_ABI}/libpaddle_light_api_shared.so
# Provides the path to the library you want to import.
)
# Specifies libraries CMake should link to your target library. You can link
# multiple libraries, such as libraries you define in this build script,
# prebuilt third-party libraries, or system libraries.
target_link_libraries(
# Specifies the target library.
Native
paddle_light_api_shared
${OpenCV_LIBS}
GLESv2
EGL
jnigraphics
${log-lib}
)
add_custom_command(
TARGET Native
POST_BUILD
COMMAND
${CMAKE_COMMAND} -E copy
${PaddleLite_DIR}/cxx/libs/${ANDROID_ABI}/libc++_shared.so
${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libc++_shared.so)
add_custom_command(
TARGET Native
POST_BUILD
COMMAND
${CMAKE_COMMAND} -E copy
${PaddleLite_DIR}/cxx/libs/${ANDROID_ABI}/libpaddle_light_api_shared.so
${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libpaddle_light_api_shared.so)
add_custom_command(
TARGET Native
POST_BUILD
COMMAND
${CMAKE_COMMAND} -E copy
${PaddleLite_DIR}/cxx/libs/${ANDROID_ABI}/libhiai.so
${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libhiai.so)
add_custom_command(
TARGET Native
POST_BUILD
COMMAND
${CMAKE_COMMAND} -E copy
${PaddleLite_DIR}/cxx/libs/${ANDROID_ABI}/libhiai_ir.so
${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libhiai_ir.so)
add_custom_command(
TARGET Native
POST_BUILD
COMMAND
${CMAKE_COMMAND} -E copy
${PaddleLite_DIR}/cxx/libs/${ANDROID_ABI}/libhiai_ir_build.so
${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libhiai_ir_build.so)

@ -1,48 +0,0 @@
//
// Created by fu on 4/25/18.
//
#pragma once
#import <vector>
#import <numeric>
#ifdef __ANDROID__
#include <android/log.h>
#define LOG_TAG "OCR_NDK"
#define LOGI(...) \
__android_log_print(ANDROID_LOG_INFO, LOG_TAG, __VA_ARGS__)
#define LOGW(...) \
__android_log_print(ANDROID_LOG_WARN, LOG_TAG, __VA_ARGS__)
#define LOGE(...) \
__android_log_print(ANDROID_LOG_ERROR, LOG_TAG, __VA_ARGS__)
#else
#include <stdio.h>
#define LOGI(format, ...) \
fprintf(stdout, "[" LOG_TAG "]" format "\n", ##__VA_ARGS__)
#define LOGW(format, ...) \
fprintf(stdout, "[" LOG_TAG "]" format "\n", ##__VA_ARGS__)
#define LOGE(format, ...) \
fprintf(stderr, "[" LOG_TAG "]Error: " format "\n", ##__VA_ARGS__)
#endif
enum RETURN_CODE {
RETURN_OK = 0
};
enum NET_TYPE{
NET_OCR = 900100,
NET_OCR_INTERNAL = 991008
};
template <typename T>
inline T product(const std::vector<T> &vec) {
if (vec.empty()){
return 0;
}
return std::accumulate(vec.begin(), vec.end(), 1, std::multiplies<T>());
}

@ -1,115 +0,0 @@
//
// Created by fujiayi on 2020/7/5.
//
#include "native.h"
#include "ocr_ppredictor.h"
#include <string>
#include <algorithm>
#include <paddle_api.h>
static paddle::lite_api::PowerMode str_to_cpu_mode(const std::string &cpu_mode);
extern "C"
JNIEXPORT jlong JNICALL
Java_com_baidu_paddle_lite_demo_ocr_OCRPredictorNative_init(JNIEnv *env, jobject thiz,
jstring j_det_model_path,
jstring j_rec_model_path,
jint j_thread_num,
jstring j_cpu_mode) {
std::string det_model_path = jstring_to_cpp_string(env, j_det_model_path);
std::string rec_model_path = jstring_to_cpp_string(env, j_rec_model_path);
int thread_num = j_thread_num;
std::string cpu_mode = jstring_to_cpp_string(env, j_cpu_mode);
ppredictor::OCR_Config conf;
conf.thread_num = thread_num;
conf.mode = str_to_cpu_mode(cpu_mode);
ppredictor::OCR_PPredictor *orc_predictor = new ppredictor::OCR_PPredictor{conf};
orc_predictor->init_from_file(det_model_path, rec_model_path);
return reinterpret_cast<jlong>(orc_predictor);
}
/**
* "LITE_POWER_HIGH" convert to paddle::lite_api::LITE_POWER_HIGH
* @param cpu_mode
* @return
*/
static paddle::lite_api::PowerMode str_to_cpu_mode(const std::string &cpu_mode) {
static std::map<std::string, paddle::lite_api::PowerMode> cpu_mode_map{
{"LITE_POWER_HIGH", paddle::lite_api::LITE_POWER_HIGH},
{"LITE_POWER_LOW", paddle::lite_api::LITE_POWER_HIGH},
{"LITE_POWER_FULL", paddle::lite_api::LITE_POWER_FULL},
{"LITE_POWER_NO_BIND", paddle::lite_api::LITE_POWER_NO_BIND},
{"LITE_POWER_RAND_HIGH", paddle::lite_api::LITE_POWER_RAND_HIGH},
{"LITE_POWER_RAND_LOW", paddle::lite_api::LITE_POWER_RAND_LOW}
};
std::string upper_key;
std::transform(cpu_mode.cbegin(), cpu_mode.cend(), upper_key.begin(), ::toupper);
auto index = cpu_mode_map.find(upper_key);
if (index == cpu_mode_map.end()) {
LOGE("cpu_mode not found %s", upper_key.c_str());
return paddle::lite_api::LITE_POWER_HIGH;
} else {
return index->second;
}
}
extern "C"
JNIEXPORT jfloatArray JNICALL
Java_com_baidu_paddle_lite_demo_ocr_OCRPredictorNative_forward(JNIEnv *env, jobject thiz,
jlong java_pointer, jfloatArray buf,
jfloatArray ddims,
jobject original_image) {
LOGI("begin to run native forward");
if (java_pointer == 0) {
LOGE("JAVA pointer is NULL");
return cpp_array_to_jfloatarray(env, nullptr, 0);
}
cv::Mat origin = bitmap_to_cv_mat(env, original_image);
if (origin.size == 0) {
LOGE("origin bitmap cannot convert to CV Mat");
return cpp_array_to_jfloatarray(env, nullptr, 0);
}
ppredictor::OCR_PPredictor *ppredictor = (ppredictor::OCR_PPredictor *) java_pointer;
std::vector<float> dims_float_arr = jfloatarray_to_float_vector(env, ddims);
std::vector<int64_t> dims_arr;
dims_arr.resize(dims_float_arr.size());
std::copy(dims_float_arr.cbegin(), dims_float_arr.cend(), dims_arr.begin());
// 这里值有点大就不调用jfloatarray_to_float_vector了
int64_t buf_len = (int64_t) env->GetArrayLength(buf);
jfloat *buf_data = env->GetFloatArrayElements(buf, JNI_FALSE);
float *data = (jfloat *) buf_data;
std::vector<ppredictor::OCRPredictResult> results = ppredictor->infer_ocr(dims_arr, data,
buf_len,
NET_OCR, origin);
LOGI("infer_ocr finished with boxes %ld", results.size());
// 这里将std::vector<ppredictor::OCRPredictResult> 序列化成 float数组传输到java层再反序列化
std::vector<float> float_arr;
for (const ppredictor::OCRPredictResult &r :results) {
float_arr.push_back(r.points.size());
float_arr.push_back(r.word_index.size());
float_arr.push_back(r.score);
for (const std::vector<int> &point : r.points) {
float_arr.push_back(point.at(0));
float_arr.push_back(point.at(1));
}
for (int index: r.word_index) {
float_arr.push_back(index);
}
}
return cpp_array_to_jfloatarray(env, float_arr.data(), float_arr.size());
}
extern "C"
JNIEXPORT void JNICALL
Java_com_baidu_paddle_lite_demo_ocr_OCRPredictorNative_release(JNIEnv *env, jobject thiz,
jlong java_pointer){
if (java_pointer == 0) {
LOGE("JAVA pointer is NULL");
return;
}
ppredictor::OCR_PPredictor *ppredictor = (ppredictor::OCR_PPredictor *) java_pointer;
delete ppredictor;
}

Some files were not shown because too many files have changed in this diff Show More

Loading…
Cancel
Save