- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
[IEEE 2018 International Conference on Soft-computing and Network Security (ICSNS) - Coimbatore, India (2018.2.14-2018.2.16)] 2018 International Conference on Soft-computing and Network Security (ICSNS) - Improving Optical Braille Recognition in Pre-processing Stage
摘要： Braille documents are captured and processed and converted into natural text using Optical Braille Recognition (OBR) system. Braille data or information which is embossed on Braille plate is most commonly captured by mobile camera or scanner into Braille image is the input to OBR. Preprocessing is the first stage performed by OBR system. The Noise introduced while Image captured using camera, mobile or scanners involve irregular lightness, low quality image due to less pixel camera, impulse noise, diverse gray-level values, angled or slanted image captured as a result of human error, deformation or warp of image document, deprived or disabled dots, appearance of unwanted dots, irregular gap between adjacent dots that represent a character. Various image enhancement techniques are applied in preprocessing stage to abolished or reduce such impulse noises. The main mission of the paper is to understand and apply a proportional learning of diverse preprocessing techniques on the Braille image and identify the best method to improve quality of Braille image.
关键词： Braille,Box filter,preprocessing,Spatial filter,Gaussian filter,Smoothing,Histogram Equalization,OCR,OBR
更新于2025-09-09 09:28:46
-
Composite filtering strategy for improving distortion invariance in object recognition
摘要： Correlation-based pattern recognition filtering methods such as the eigen-extended maximum average correlation height (EEMACH) filter is considered an effective tool in object recognition applications. However, these approaches require exclusive training for all possible distortions including in-plane as well as out-of-plane rotation, scale and illumination variations. The overall training process is exhaustive and requires training of filter banks to handle specific types of distortion separately. To overcome the aforementioned limitations, the authors propose a new difference of Gaussian (DoG)-based logarithmically preprocessed EEMACH filter which can manage multiple distortions in a single training instance while ensuring inherent control over illumination variations. The DoG-based logarithmic treatment exploits inherent capabilities of logarithmic preprocessing to manage scale and in-plane rotations. By reducing the number of classifier instances to one third, it not only reduces the computation complexity of the process to 33%, but also enhances the object recognition performance. The cumulative improvement is up to 2.73% in case of rotations and 10.8% in case of scaling by incorporating reinforced edges due to DoG operation. The resultant filter displays significantly enhanced recognition performance leading to a higher percentage of correct machine decisions, especially when an input scene contains multiple distortions.
关键词： Correlation-based pattern recognition,logarithmic preprocessing,difference of Gaussian,distortion invariance,object recognition,EEMACH filter
更新于2025-09-04 15:30:14
-
[ACM Press the 2018 International Conference - Hong Kong, Hong Kong (2018.02.24-2018.02.26)] Proceedings of the 2018 International Conference on Image and Graphics Processing - ICIGP 2018 - Image Change Detection Using Comprehensive Preprocessing and Spectral Log-Demons Algorithm
摘要： This paper proposes a universal framework for image change detection for daily basic use. Overall, the challenge lies on the interference factors such as noise captured during shooting, illumination variation in time-series images and rotation angle of the view. To overcome these problems, we introduce the Comprehensive Preprocessing and Spectral Log-Demons Algorithm (CPP-SLD). We begin our methodology with Comprehensive Preprocessing methods, aiming to eliminate the irrelevant factors. We analyze the sparseness property of gradient distribution to correct the bias. In addition, we make use of the symmetric distribution of the radial gradient to correct the vignetting. Besides, Wiener filtering and Gaussian filtering are both used to denoise and smooth the images. Finally, the Spectral Log-Demons algorithm is applied to register. Then by subtraction we detect the change. Our framework can achieve a widespread success under different circumstances. Moreover, we compare our method with 3 ohthers, and by qualitative and quantitative analysis we expound the validity and robustness of our method.
关键词： comprehensive preprocessing,Spectral Log-Demons Algorithm,correction,Image change detection,filter.
更新于2025-09-04 15:30:14