The Research And Application Of Uneven Illumination Of Text Image Correction Algorithm

With the rapid development of information technology, the use of computer processing text messages has become an inevitable trend. Traditional paper-based carrier is the text, the paper text into computer text processing communal used are: image acquisition, image preprocessing, character recognition. Due to the impact of the shooting equipment device itself and the different shooting conditions and the difference of the shooting image itself absorption and reflection of light, making the image uneven illumination is widespread, also make the uneven illumination image correction to improve text recognition becomes more difficult and important,In this paper, put the uneven illumination text image as the object of study, and put the quality improvement of the image, the complete separation of target and background point as the ideas of study, and put the improvement of the recognition rate as the ultimate purpose, and put three types of quality correction method like image enhancement, image Binarization, image denoising as the processing methods, and expand analysis and discussion on to the content.On the basis of the above analysis, this paper puts forward the uneven illumination of text image correction method based on area. First, pull the difference between the background point and the target point difference though the method of3D enhancement based on neighborhood. Secondly, use the projection algorithm to realize the regional classification after the analysis of the characteristics of the text image, and deal the white areas and the even illumination area with the global correction after the realization of regional classification. For the shaded area, deal with the local correction processing combined with the characteristic parameter for complete separation of the background points and the target points. Finally, adopt the denoising method based on the neighborhood with the analysis of the noise characteristic to correct the misclassification point.The experimental results show that this algorithm can effectively separate the target pixel and background pixel of the uneven illumination of text image, and the visual rendering after processing can be used for reading and improve the recognition rate of OCR system. Finally, the article summarizes the advantages and disadvantages of the proposed algorithm and the future direction of development.