Document-detection-and-Skew-correction

At first I'm doing image alignment then doing document detection,Here i'm using google vision api to detect document and parse content from samples with 100% accuracy.Then parse specific content from invoice samples.

Dependency

python 3.7
pip install tensorflow=2.0.0
pip install pandas
pip install numpy

alined samples,Detect Document and Parse content

Run above command
image_alignmnet-2 (1).ipynb

Result:
1)non_aligmnet_sample

2)detected sample

3)alignment_sample

Parse specific content

Run above comment:
parse_shop_name.ipynb
sample:

Result:
Like parse shop name from invoice samples

References