Marmelator/image-classification-service
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 29 Wiki Activity

add new pyinfra, add optional image classifcation under key dataCV if figure metadata is present on storage

Browse Source
This commit is contained in:
Julius Unverfehrt 2022-08-12 13:37:48 +02:00
parent bae25bedbd
commit 2b2da1b60c
5 changed files with 11 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
image_prediction/default_objects.py
View File

@ -30,12 +30,6 @@ def get_image_classifier(model_loader, model_identifier):
return ImageClassifier(Classifier(EstimatorAdapter(model), ProbabilityMapper(classes))) return ImageClassifier(Classifier(EstimatorAdapter(model), ProbabilityMapper(classes)))
# def get_extractor(**kwargs):
# image_extractor = ParsablePDFImageExtractor(**kwargs)
#
# return image_extractor
def get_dispatched_extract(**kwargs): def get_dispatched_extract(**kwargs):
image_extractor = ParsablePDFImageExtractor(**kwargs) image_extractor = ParsablePDFImageExtractor(**kwargs)

6
image_prediction/formatter/formatter.py
View File

@ -21,9 +21,9 @@ class Formatter(Transformer):
def format_image_plus(image: ImagePlus) -> ImageMetadataPair: def format_image_plus(image: ImagePlus) -> ImageMetadataPair:
enum_metadata = { enum_metadata = {
Info.PAGE_WIDTH: image.info.pageInfo.width, Info.PAGE_WIDTH: image.info.pageInfo.pageWidth,
Info.PAGE_HEIGHT: image.info.pageInfo.height, Info.PAGE_HEIGHT: image.info.pageInfo.pageHeight,
Info.PAGE_IDX: image.info.pageInfo.number, Info.PAGE_IDX: image.info.pageInfo.pageNumber,
Info.ALPHA: image.info.alpha, Info.ALPHA: image.info.alpha,
Info.WIDTH: image.info.boundingBox.width, Info.WIDTH: image.info.boundingBox.width,
Info.HEIGHT: image.info.boundingBox.height, Info.HEIGHT: image.info.boundingBox.height,

2
incl/pdf2image

@ -1 +1 @@
Subproject commit 699568875683ba727ec9759c8bea85e0d3e1d369 Subproject commit fee87964cb7da0ea0c19410ca418849744474302

2
incl/pyinfra

@ -1 +1 @@
Subproject commit 0f6512df5423df98d334f5735170cd1f7642998a Subproject commit be82114f8302ffedecf950c6ca9fecf01ece5573

12
src/serve.py
View File

@ -31,17 +31,17 @@ def process_request(request_message):
object_bytes = storage.get_object(PYINFRA_CONFIG.storage_bucket, f"{dossier_id}/{file_id}.{target_file_extension}") object_bytes = storage.get_object(PYINFRA_CONFIG.storage_bucket, f"{dossier_id}/{file_id}.{target_file_extension}")
object_bytes = gzip.decompress(object_bytes) object_bytes = gzip.decompress(object_bytes)
try: # TODO: add figure detection file target to request message to avoid this if storage.exists(PYINFRA_CONFIG.storage_bucket, f"{dossier_id}/{file_id}.FIGURE.json.gz"):
metadata_bytes = storage.get_object(PYINFRA_CONFIG.storage_bucket, f"{dossier_id}/{file_id}.FIGURE.json.gz") metadata_bytes = storage.get_object(PYINFRA_CONFIG.storage_bucket, f"{dossier_id}/{file_id}.FIGURE.json.gz")
metadata_bytes = gzip.decompress(metadata_bytes) metadata_bytes = gzip.decompress(metadata_bytes)
metadata_per_image = json.load(io.BytesIO(metadata_bytes))["data"] metadata_per_image = json.load(io.BytesIO(metadata_bytes))["data"]
logger.info("Metadata acquired") classifications_cv = list(pipeline(pdf=object_bytes, metadata_per_image=metadata_per_image))
except: else:
metadata_per_image = None classifications_cv = []
classifications = list(pipeline(pdf=object_bytes, metadata_per_image=metadata_per_image)) classifications = list(pipeline(pdf=object_bytes))
result = {**request_message, "data": classifications} result = {**request_message, "data": classifications, "dataCV": classifications_cv}
response_file_extension = request_message["responseFileExtension"] response_file_extension = request_message["responseFileExtension"]
storage_bytes = gzip.compress(json.dumps(result).encode("utf-8")) storage_bytes = gzip.compress(json.dumps(result).encode("utf-8"))