Research On Reading Algorithm Of Printed Quantum Dot Information Under Non-correction Condition

Halftone information hiding anti-counterfeiting technology is a mainstream technology in the field of printing anti-counterfeiting.This technology has its undeniable advantages and requires very low cost,thus it has been widely applied in various fields.Additionally,the printed products using this technology possess characteristics that are difficult to replicate.However,in the actual printing process using this method,there are issues such as inkjetting,dot misalignment,and even printing defects.During the transportation of printed matter such as trademarks,there are problems such as fading,soiling,and even damage.Moreover,during the reading process,factors such as the shooting environment,shooting distance,azimuth angle,and jitter can affect the recognition,often leading to certain degree of information omission or even errors.Therefore,under the condition of ensuring the speed of image recognition algorithms,it is of great research value to reliably extract the printed halftone image information.Considering the advantages of halftone information hiding and its significant role in the field of anti-counterfeiting,this thesis first conducts research on the information extraction techniques for popular images embedded with hidden information.It proposes a fast information extraction algorithm specifically designed for halftone printed halftone dot images.Without the need for geometric correction,the algorithm follows the following steps: Preprocess and binarize the folded and soiled halftone dot image to obtain a standardized binary image;Perform center calibration on the binary image proportionally and crop the non-full-bleed image edges or extend the boundaries of the full-bleed image to prevent abnormal termination during subsequent reading processes;Extract the central region of the non-soiled part of the image as the prior region and generate a sequence by accumulating the pixels in that region.The optimized result of the estimated minimum fill ratio is used as the main direction offset angle of the halftone dot grid;Perform differential processing on the accumulated sum sequence corresponding to the offset angle to record the pixel values at positive and negative alternations.Adjust the pixel value sequence within an error tolerance based on the second-order differential result of the accumulated sum sequence;Apply the same method to the sub-direction of the prior region of the full-bleed image or use the minimum spacing between adjacent reference lines of the main direction in the prior region of the non-full-bleed image as a threshold for iterative recursive determination of each printed halftone dot.This generates a complete prior region halftone dot recognition matrix;Generate a dynamic vector based on the initial value of neighboring printed halftone dot distances and update it during the search process.Perform recursive search on the four neighboring areas of the prior region in the image based on the dynamic vector,generating local halftone dot recognition matrices in the order of search;Finally,use the edge rows and columns of the generated recognition matrices as initial conditions and continue searching the remaining areas of the image.Transpose and expand the local matrices based on the spatial distribution of the halftone dots,and then concatenate them to form a complete recognition dot matrix.By following this algorithm,the proposed approach enables fast information extraction from halftone printed images with hidden information,without the need for geometric correction.The effectiveness and reliability of the recognition process were compared for images with different sizes,total number of printed halftone dots,and levels of distortion and contamination.The proposed algorithm not only enables direct extraction of the printed halftone dot information from the images but also works effectively for images with trapezoidal,pillow-shaped,spherical,and cylindrical distortions.This method can be well applied in the field of copyright protection,enhancing the level of information value-added services and promoting the wider application of halftone anti-counterfeiting technology.