Text detection method, device and equipment and storage medium

Abstract

The invention provides a text detection method, device and equipment and a storage medium. The method comprises the steps: obtaining an initial feature image of a to-be-recognized text image; performing feature extraction on the initial feature image through a preset model containing a plurality of channels to obtain a target feature image for the plurality of channels, and representing a text region probability graph by feature information extracted from a first channel of the plurality of channels in the target feature image, the total feature information extracted by all the other channelsexcept the first channel in the multiple channels representing the total coding values of the pixel points corresponding to the text areas, the total coding values of the pixel points corresponding tothe same text area being the same, and the total coding values of the pixel points between different text areas being different; and determining a feature value corresponding to each pixel point in the to-be-identified text image based on the probability value of each pixel point represented by the text region probability graph and the total coded value extracted based on all other channels, so as to determine a plurality of text regions of the to-be-identified text image.

Description

technical field [0001] The present application relates to image processing technology, in particular to a text detection method, device, equipment and storage medium. Background technique [0002] Text detection has a wide range of applications and is a pre-step for many computer vision tasks. Its main purpose is to locate the position of text lines or characters in the image. The precise positioning of text is both very important and challenging, because compared with general object detection. In other words, text has the characteristics of multi-direction, irregular shape, extreme aspect ratio, font, color, background, etc. Therefore, algorithms that are more successful in general object detection cannot be directly transferred to text detection. Although, with the rise of deep learning technology, some text detection methods have been proposed. However, for denser texts, such as arithmetic exercises for primary school students, there are 100 text areas on an image. At thi...

AI Technical Summary

Problems solved by technology

[0002] Text detection has a wide range of applications and is a pre-step for many computer vision tasks. Its main purpose is to locate the position of text lines or characters in the image. The precise positioning of text is both very important and challenging, because compared with general object detection. In other words, text has the characteristics of multi-direction, irregular shape, extreme aspect ratio, font, color, background, etc. Therefore, algorithms that are more successful in general object detection cannot be directly transferred to text detection.

Although, with the rise of deep learning technology, some text detection methods have been proposed. However, for denser texts, such as arithmetic exercises for primary school students, there are 100 text areas on an image. At this time, the processing of existing methods The rate will decrease linearly with the increase of the number of text boxes. Therefore, it cannot meet the needs of actual application scenarios and reduces the user experience.

Images