Enable orientation detection. Document orientation will be returned in DocumentQualityAnalyzerResult.orientation.
Default is false
Inspect small text (typically less than 12 pixels height) when analyzing the document quality. If true, processing will be slower, but when the quality issue with a document is the low resolution of the text, then the issue will be caught.
Default is false
Maximum image size in pixels, if image is bigger, it will be resized.
Default is 2000
At most this fraction of the image will be processed, range is from 0 to 1. Decreasing this value below 1.0 can lead to faster processing but may result in less accurate prediction, as only a subset of the image tiles will be analyzed.
Default is 1
If estimated number of symbols is less than this value, return that document is not found.
Default is 20
If this number of symbols is found and minProcessedFraction of the image is processed, the processing stops.
Default is 100
At least this fraction of the image will be processed, range is from 0 to 1. Decreasing this value below 1.0 can lead to faster processing but may result in less accurate prediction, as only a subset of the image tiles will be analyzed.
Default is 1
If detectOrientation is True, minNumberOfRequiredSymbols is found, and orientationConfidence is bigger than this threshold, the processing stops. If the maxProcessedFraction part of the image is processed and orientationConfidence is lower than the threshold, the returned orientation will be 0.
Default is 0
quality levels.
Quality thresholds to separate quality levels.
If true, will return the quality heatmap as a part of the result. In this case, the entire image will be processed and the minProcessedFraction will be ignored. Useful for debugging situations where the returned quality doesn't appear to match the perceived quality.
Default is false
Image will be processed in tiles of this size; will be ignored if image is small.
Default is 300
Document Quality Analyzer configuration.