Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.abbyy.com/llms.txt

Use this file to discover all available pages before exploring further.

When the program matches an Extraction Rules activity, it attempts to locate objects that correspond to the activity’s elements, as well as evaluate the accuracy of the results. During matching, elements are searched for sequentially in the order in which they are described in the Search Elements tree (descending order). Each element in the specified search area can be matched to several image objects that fit the element description. The program considers the various objects in the form of hypotheses regarding the location of elements. A hypothesis is an assumption that the detected object suits a particular element, i.e. satisfies the specified element properties and search conditions.
The element search area may contain several objects that can be matched to the element. In this case, several element hypotheses are formulated.
The primary characteristic of a hypothesis is its hypothesis quality, which is an evaluation of how well the detected object matches the given element (its properties and search conditions). Hypothesis quality ranges from 0 to 1. Hypothesis quality for a group element is calculated by multiplying together the quality values of all its subelement hypotheses. A quality value of 1 means that the detected object fits the description exactly (or 100%). If the value is not 0, then matching will also include a step where field regions for the Extraction Rules activity are formed.