Universal detection algorithm for optical printing text and scene text

A technology for optical printing and general detection, which is applied in the field of general detection algorithms for optically printed text and scene text, which can solve the problems of affecting the effect, serious time-consuming, complicated process, etc., to improve integrity, improve accuracy, and solve low efficiency effective effect

Pending Publication Date: 2021-05-28
浙江康旭科技有限公司
5 Cites 0 Cited by

AI-Extracted Technical Summary

Problems solved by technology

[0005] First of all, the existing technology is relatively mature for text detection in a single direction, such as horizontal and vertical directions, but it cannot achieve an accurate detection of scene texts of various shapes (curved, vertical, multi-directional), based on bank For bill text recognition, in some business scenarios, it is necessary to detect the s...
View more

Method used

S3, distinguish the type of different chapters according to bank scene demand, and use OpenCV technology, remove red seal, use OpenCV technology to remove the specific implementation method of red seal is, color map is carried out separation channel, extracts the red channel of picture, by Set the threshold range to remove the r...
View more

Abstract

The invention discloses a general detection algorithm for an optical printing text and a scene text, and the algorithm comprises the following steps: S1, obtaining a to-be-detected picture, S2, detecting a target text region in the to-be-detected picture, S3, determining a detection model of a target text corresponding to the target text region according to a preset corresponding relation between a text region and a text detection model, and S4, obtaining a target text detection box by adopting the target text detection model. In the invention, the DBnet network algorithm model can test the performance of the detection algorithm in the text recognition under the condition of not changing the source code of the text recognition model, and adjust the text detection algorithm based on the test result, so that the accuracy of the text detection algorithm can be improved, the accuracy of the whole text recognition model is also improved, the text detection integrity can be improved, and the problem that in the prior art, various text types in a complex layout cannot be completely detected by adopting a single process is solved.

Application Domain

Character and pattern recognition

Technology Topic

EngineeringText detection +4

Image

  • Universal detection algorithm for optical printing text and scene text
  • Universal detection algorithm for optical printing text and scene text
  • Universal detection algorithm for optical printing text and scene text

Examples

  • Experimental program(1)
  • Effect test(1)

Example Embodiment

[0066]Example one
[0067]SeeFigure 1-3The present invention provides a technical solution: an universal detection algorithm for optical printing text and scene text, including the following steps:
[0068]S1, get the picture to be detected;
[0069]S2, detect the target text area in the picture to be detected;
[0070]S3, determine the detection model of the target text area corresponding to the target text according to the correspondence between the pre-set text area and the text detection model;
[0071]S4, using the target text detection model to obtain the target text detection box;
[0072]The target text area includes a header text area, a seal text area and a common text area;
[0073]The correspondence between the text area and the text detection model includes at least one of: using the DBNET network algorithm model for text detection and using the OpenCV tool according to the threshold;
[0074]The establishment of the DBNET network algorithm model includes the following steps:
[0075]S1, training data set preparation;
[0076]S2, use the DBNET network text detection algorithm for migration learning;
[0077]Text detection includes the following steps:
[0078]S1, collect pictures that need to be detected;
[0079]S2, using the open source pre-training model of the DBNET network algorithm for pre-training, a fixed threshold segmentation map and an adaptive threshold segmentation are generated;
[0080]S3, combined with fixed threshold segmentation map and adaptive segmentation graph generation bisarrative diagram;
[0081]S4, get the text detection box to be detected, generate quadrilateral and polygonal detection box, training data set is a data set with Chinese, English street view, including 30,000 Chinese bidding data set and 10,000 English bidding data set;
[0082]Training data concentration horizontal entry and vertical entry position labeled 4 points, curved text marked as polygons and bending text Note, DBNET Network Algorithm Model Digest Polygon;
[0083]The formula of the image multilateral clipping algorithm is:
[0084]
[0085]among them:
[0086]D is the contraction offset of the polygon, and the perimeter of the polygon, the area of ​​the polygon, R is the formula of the loss function used in the contraction factor, the DBNET network algorithm model::
[0087]L = Ls+ α × Lb+ β × Lt;

PUM

no PUM

Description & Claims & Application Information

We can also present the details of the Description, Claims and Application information to help users get a comprehensive understanding of the technical details of the patent, such as background art, summary of invention, brief description of drawings, description of embodiments, and other original content. On the other hand, users can also determine the specific scope of protection of the technology through the list of claims; as well as understand the changes in the life cycle of the technology with the presentation of the patent timeline. Login to view more.

Similar technology patents

Adaptive fault detection method for airplane rotation actuator driving device based on deep learning

InactiveCN104914851Aimprove accuracyReduce the false alarm rate of detection
Owner:BEIHANG UNIV

Communication terminal apparatus and communication system

InactiveUS20060003776A1fluctuation of be improveimprove accuracy
Owner:SONY ERICSSON MOBILE COMM JAPAN INC

Classification and recommendation of technical efficacy words

  • improve accuracy
  • improve integrity

Golf club head with adjustable vibration-absorbing capacity

InactiveUS20050277485A1improve grip comfortimprove accuracy
Owner:FUSHENG IND CO LTD

Stent delivery system with securement and deployment accuracy

ActiveUS7473271B2improve accuracyreduces occurrence and/or severity
Owner:BOSTON SCI SCIMED INC

Method for improving an HS-DSCH transport format allocation

InactiveUS20060089104A1improve accuracyincrease benefit
Owner:NOKIA SOLUTIONS & NETWORKS OY

Catheter systems

ActiveUS20120059255A1increase selectivityimprove accuracy
Owner:ST JUDE MEDICAL ATRIAL FIBRILLATION DIV

Gaming Machine And Gaming System Using Chips

ActiveUS20090075725A1improve accuracy
Owner:UNIVERSAL ENTERTAINMENT CORP

CT integrated respiratory monitor

InactiveUS7182083B2improve integrityeasy to train
Owner:KONINKLIJKE PHILIPS ELECTRONICS NV

Bone anchors for orthopedic applications

InactiveUS20100082072A1improve integrity
Owner:WARSAW ORTHOPEDIC INC

GPS/INS vehicle attitude system

InactiveUS20050137799A1improve integritylow-cost
Owner:THE BOARD OF TRUSTEES OF THE LELAND STANFORD JUNIOR UNIV

Vehicle body structure

ActiveUS20050212333A1minimize local deformationimprove integrity
Owner:NISSAN MOTOR CO LTD

Local airport monitoring method and system based on multi-receiver

InactiveCN101419275AImprove monitoring and judgment abilityImprove integrity
Owner:BEIHANG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products