Image detection method and device, storage medium and equipment

Abstract

The invention discloses an image detection method and device, a storage medium and equipment, and the method comprises the steps: firstly, carrying out the detection of an obtained target image, obtaining a binary image containing a text line position and a binary image containing a single word position, then carrying out the marking of a text line in the binary image containing the text line position, and carrying out the marking of the single word position; obtaining an identifier corresponding to each text line; determining the coordinate value of each pixel point in the position of each text line, and calculating the coordinate value of the pixel point in the central position of the position of each single character in the binary image containing the position of the single character; next, the coordinate value of the pixel point at the central position of the position of each single character is matched with the coordinate value of each pixel point in the position of each text line, and the text line identifier to which each single character belongs is determined, so that each single character can be filled into the text line of the corresponding identifier; accurate matching of the position of the text line and the position of the single character is achieved, and therefore the accuracy of the detection result and the detection efficiency are improved.

Description

technical field [0001] The present application relates to the technical field of image processing, and in particular to an image detection method, device, storage medium and equipment. Background technique [0002] With the continuous breakthrough and development of artificial intelligence (AI) technology. The demand for optical character recognition (Optical Character Recognition, OCR for short) recognition technology is gradually increasing. Among them, multi-granularity OCR detection refers to the simultaneous detection of text lines and single characters in an image, which is equivalent to simultaneously detecting text line position information and single character position information. Because some business scenarios require both text line information and single word information. Therefore, how to accurately match and integrate the position of the text line and the position of the word is an important issue in related fields at present. [0003] At present, when perf...

AI Technical Summary

Problems solved by technology

However, for the image scene with dense text, when matching the position of the text line and the position of the single word, a large number of such as figure 1 The shown box calculates the IOU, which leads to a slow matching speed. This is because for polygons, it is necessary to use a toolkit such as Polygon to calculate the IOU of the two boxes, which makes the calculation complexity (for the number of single characters and the text The product of the number of rows) is higher

Moreover, for irregular shapes (such as concave polygons, etc.), considering the calculation speed, it will be converted into the minimum circumscribed rotation rectangle (minAreaRect) for calculation when calculating the IOU, which is suitable for tilt, irregular shape, For the location box with a containment relationship, the calculated IOU result is not accurate enough, which leads to poor final matching effect, and it is impossible to obtain a more accurate text detection result.

Images