Original Binary Image Research Articles

Edge Detection Techniques Using Character Segmentation for OCR of Devanagari Text

An image processing technique called edge detection is used to find and identify the edges, curves, and objects in digital images. Edge detection includes various mathematical calculations that focus on locating the edges and curves where a digital image's brightness abruptly shifts or has discontinuities. In the fields of feature detection and feature extraction in image processing and computer vision, edge detection is a fundamental tool. The application of an edge detector results in the development of a set of connected curves that represent surface marks, character boundaries, and discontinuities in surface orientation. Character boundaries are described via edges. The majority of a feature is described by its edges. A group of pixels called an edge is used to describe an area where abrupt variations in intensity occur. Because it removes items of interest for actions like description, segmentation is a vital stage in an image recognition system. This paper focuses on edge detection techniques that can be implemented on characters for effective character recognition. Devanagari text includes curvy edges which need to be detected for proper recognition hence different edge detector operators are implemented to find the most effective amongst them to acquire clear edges of the text. There are several methods for segmenting images, including segmentation based on artificial neural networks, clustering, edges, regions, and thresholds. In this paper, edge-based segmentation is used. The Canny, Laplacian, Sobel, Prewitt, and Robert operators are used to detect edges, and the generated image is compared to the original binary image. The pre-processing phase includes the binarization step.

A Novel Steganography-Based Pattern for Print Matter Anti-Counterfeiting by Smartphone Cameras.

Print matter authentication based on anti-counterfeiting techniques has received continuously increasing concern from academia and industry. However, the existing printing anti-counterfeiting solutions often have the defects of poor identification experience, high cost, or weak anti-counterfeiting ability, and cannot achieve pre-sale anti-counterfeiting. Therefore, a novel steganography-based pattern for print matter anti-counterfeiting by smartphone cameras is proposed in this study. Firstly, every pixel in the original binary message image (such as QR code) is replaced by a square pixel block with the same binary gray value of 0 or 255 (the first-level expansion). Secondly, the obtained image is encrypted based on the logistic chaotic sequence, and then scrambled by Arnold transform. Lastly, once again every pixel in the generated image is replaced with a square pixel block (the second-level expansion), the size and gray value of which can be set to control the semi-fragile ability to distinguish an originally printed pattern from its illegitimate copy. If the message extracted from the printed pattern through the inverse procedure is complete enough to decode and read, the pattern is assumed to be an original print. Experimental results verify the advancement and effectiveness of the proposed scheme in distinguishing the copied pattern.

Reversible data hiding in binary images by flipping pattern pair with opposite center pixel

In this paper, a novel RDH scheme by flipping pattern pairs with opposite center pixel (PPOCPs) in binary images is proposed, aiming at decreasing the distortion while increasing the embedding payload. First, 25 patterns in the 3×3 block are designed which construct the PPOCPs according to the distance level providing a guarantee for reversibility. Then, a balanced score is designed which considers both visual distortion and embedding payload to select the optimal PPOCP, and the secret messages are embedded in the optimal PPOCP. For the receiver, the secret messages can be extracted precisely and the original binary image can be recovered by scanning the optimal PPOCP. PPOCP is a novel RDH model which fully considers the visual distortion caused by flipping pixels. Experimental results demonstrate the feasibility of the proposed RDH method for binary images, and the visual quality is satisfactory under high embedding payload and smallest pure flipping rate.

A compressed secret image sharing method with shadow image verification capability.

Secret image sharing (SIS) is an important research direction in information hiding and data security transmission. Since the generated shadow images (shares) are always noise-like, it is difficult to distinguish the fake share from the unauthorized participant before recovery. Even more serious is that an attacker with a fake share can easily collect shares of other honest participants. As a result, it is significant to verify the shares, before being taken out for recovery. Based on two mainstream methods of SIS, such as polynomial-based SIS and visual secret sharing(VSS), this paper proposed a novel compressed SIS with the ability of shadow image verification. Considering that the randomness of the sharing phase of polynomial-based SIS can be utilized, one out of shares of (2, 2)-threshold random-grid VSS is embedded into all shares of polynomial-based SIS by a XOR operation as the verification information, while the other binary share is private for verification. Before recovery, each participant must extract the binary share from the grayscale share to perform XOR operation with the private share, and the original binary image can be recovered only with the true share. The proposed scheme also has the characteristics of shadow image verification, pixel compression, loss tolerance and lossless recovery. Through experiments and comparative analysis of related research results, the effectiveness and advantages of the method are verified.

A New MRF-Based Lossy Compression for Encrypted Binary Images

Although there exist many researches on the compression of original non-encrypted binary images, few approaches focus on the compression of encrypted binary images. As binary images like contract, signature, halftone images are still used widely in practice, how to compress efficiently encrypted binary images in a lossy way deserves further exploration. To this end, this paper develops a lossy compression scheme for encrypted binary images by exploiting the Markov random field (MRF) model. Considering that the third-party in scenarios of cloud or distributed computing cannot access to the encryption key, we develop the concatenated down-sampling and LDPC-based encoding to perform the compression, in which four different down-sampling methods are designed to facilitate improving the quality of reconstructed image. In reconstruction, we first formulate the lossy reconstruction from the encrypted and compressed binary image as an optimization problem, and then build a joint factor graph involving the LDPC-decoding, decryption, and MRF to solve this optimization problem, in which the MRF is exploited to well infer pixels discarded in the down-sampling process. By adapting the sum-product algorithm (SPA) to the constructed joint factor graph for lossy reconstruction (JFG-LR) and running the adapted SPA on the JFG-LR, we thus recover the original binary image in a lossy way. By integrating the stream-cipher-based encryption, the down-sampling and LDPC-based compression, and the JFG-LR-involved reconstruction, we thus propose a new lossy compression scheme for encrypted binary images. Experimental results show that the proposed scheme achieves desirable compression efficiency, which is comparable to or even better than that of the JBIG2 with the original unencrypted binary image as input.

Stereo image zero-watermarking algorithm based on ternary polar harmonic Fourier moments and chaotic mapping

In recent years, stereo images have attracted extensive attention because of their strong immersion, and the corresponding copyright protection of stereo images is becoming increasingly urgent. At present, most of the watermarking algorithms for image copyright protection are aimed at planar images, and there are few watermarking algorithms for stereo images. Moreover, most of the existing stereo image watermarking algorithms are not very good at reflecting and retaining the specific relationship between the left and right views of the stereo images, which will inevitably affect the robustness of the algorithm. In this paper, ternary polar harmonic Fourier moments (TPHFM) for stereo images is proposed and based on this and chaotic mapping, a robust stereo image zero-watermarking algorithm is presented. First, mixed linear-nonlinear coupled map lattice is used to make the original binary logo image chaotic. Next, the TPHFM of the original stereo image is computed and accurate moments for the zero-watermarking algorithm are selected. Then, the binary feature image is constructed using the accurate moments, and scrambled using Sine mapping and Cosine mapping. Finally, a zero-watermark image is generated using the exclusive-or on the scrambled binary feature image and the chaotic binary logo image. Experimental results show that the proposed zero-watermarking algorithm is excellently robust against common image processing attacks and geometric attacks, and is superior to ternary radial harmonic Fourier moments (TRHFM)-based algorithm and other zero-watermarking algorithms.

Watershed Algorithm Based Image Segmentation Using Moment Preserving Bilevel Thresholding and Extended-Maxima Transform for Kernels

Watershed 변환은 수학적 형태학 분야로부터 나온 인기 있는 세분화 방법이다. 쌀, 밀, 옥수수는 전통적인 watershed 알고리듬에 의해 과세분화되어진다. 그러므로 이 논문은 확장된 최대 변환을 위해 이중 레벨의 임계치를 순간 유지하는 임계치값의 자동 선택에 의해 향상된 watershed 세분화 알고리듬을 제안한다. 첫 번째로 변형된 거리 영상은 자동 선택된 임계치를 위해 이중 레벨의 임계치를 순간 유지하여 진행되어 진다. 선택된 임계치는 확장된 최대 변환을 위해 사용되고 watershed 변환은 마무리 된다. watershed 등선 라인은 원 이진영상으로 중첩되어진다. 실험적 결과는 이런 진행과정들은 더 크게 확장하여 과세분화된 문제점을 해결 하는 것을 보인다.

A Simple Hole Filling Algorithm for Binary Cell Images

Filling holes in binary images is often required during medical image processing and analysis. However, traditional hole filling algorithms for medical images expose some disadvantages such as possible edge degradations and relatively low efficiency. To overcome such limits, a hole filling algorithm for binary cell images based on largest connected region extraction is proposed in this paper. Since there are less pixels for foreground areas in usual binary cell images, the holes in the binary images can be simply filled by extracting & filling the largest connected region in correspondent negative images together with a following OR operation between the obtained result image and the original binary image. Such an algorithm based on largest connected region extraction nearly has no effect on edge information of the ROI(region of interested) with high accuracy and efficiency. The final experiments show its validity and advantage over the traditional morphology methods for hole filling in binary cell images. Thus the presented method is of a considerable reference value for binary image processing.

An optimization for binarization methods by removing binary artifacts

In this article, we introduce a novel technique to remove binary artifacts. Given a gray-intensity image and its corresponding binary image, our method detects and remove connected components that are more likely to be background pixels. With this aim, our method constructs an auxiliary image by the minimum-error-rate threshold and, then, computes the ratio of intersection between the connected components of the original binary image and the connected components of the auxiliary image. Connected components with high ratio are considered true connected components while the rest are removed from the output. We tested our method in binarization methods for historical documents (handwritten and printed). Our results are favorable and indicate that our method can improve the outputs from diverse binarization methods. In particular, a high improvement was observed for printed documents. Our method is easy to implement, has a moderate computational cost, and has two parameters whose model interpretation allows an easy empirical selection.

Novel Binary Document Image Watermarking Exploitingthe Features of Double Domains

This paper presents a double domain watermarking scheme for binary document image. The watermark patterns are generated as the discrete cosine transform (DCT) domain signals, and then perceptually shaped through weighting its components in the spatial domain with the perceptual masks. The perceptual masks are obtained by applying a psychovisual model in the spatial domain and simultaneously considering the impact of the distance between two flappable pixels. Watermark extraction can be operated by applying a correlation based detector in either of the two domains. The approach exploits the information of the two domains for embedding, but avoids transforming the original binary image directly, which does not cause unacceptable loss of watermark. The experimental results show that the presented approach performs very well both in invisibility and robustness.

A High-Capacity Lossless Data Hiding Scheme for Binary Images

Many previous binary image lossless data hiding schemes often cause serious perceptible distortion and have poor visual quality of stego-images. To obtain better invisibility and higher hiding capacity, a novel reversible data hiding scheme for binary images is proposed by combining pattern substitution and human visual perception. Firstly, an improved binary image boundary detection method is presented by avoiding successive boundary pixel flipping. Secondly, secret messages are embedded into binary image boundaries by employing interlaced block partition. Finally, to achieve lossless recovery of original binary image, the side-information is lossless compressed by using the Golomb Code and the occurrence frequency of each pattern. Extensive experiments show that the proposed scheme is effective, and obtain at least 3.6 dB higher PSNR values of stego-image and 92.5% higher embedding capacity compared with Ho et al’s method.

Application of two-layer Quantum Dot array in image processing

In this paper a generalised two layer quantum dot model is proposed for image processing, in which two layers are coupled electrically. We found that the two layer quantum dot efficiently behaves for the application of image processing. One of the two given images is set to the input of the first layer and initial state of the second layer. The execution result is obtained from the output of the second layer. The original binary image is inputted to the first layer input and set as the initial state of the second layer.

Encryption of a volume hologram by complementary input image and binary amplitude mask

A novel volume-hologram encryption system is proposed, in which the stored binary image can be restored only by the correct random binary amplitude mask. In our system the encryption is done by overlapping two holograms in the same volume of the crystal; one is the hologram of the original binary image and the other is that of the complementary image. The principle of the system is explained with schematic diagrams and then the experimental data is shown.

Fast connected-component labelling in three-dimensional binary images based on iterative recursion

We propose two new methods to label connected components based on iterative recursion: one directly labels an original binary image while the other labels the boundary voxels followed by one-pass labelling of non-boundary object voxels. The novelty of the proposed methods is a fast labelling of large datasets without stack overflow and a flexible trade-off between speed and memory. For each iterative recursion: (1) the original volume is scanned in the raster order and an initially unlabelled object voxel v is selected, (2) a sub-volume with a user-defined size is formed around the selected voxel v, (3) within this sub-volume all object voxels 26-connected to v are labelled using iterations; and (4) subsequent iterative recursions are initiated from those border object voxels of the sub-volume that are 26-connected to v. Our experiments show that the time-memory trade-off is that the decrease in the execution time by one-third requires the increase in memory size by 3 orders. This trade-off is controlled by the user by changing the size of the sub-volume. Experiments on large three-dimensional brain phantom datasets (362 × 432 × 362 voxels of 56 MB (megabytes)) show that the proposed methods are three times faster on the average (with the maximum speedup of 10) than the existing iterative methods based on label equivalences with less than 1 MB memory consumption. Moreover, our algorithms are applicable to any dimensional data and are less dependant on the geometric complexity of connected components.

Image thinning using pulse coupled neural network

PCNN-pulse coupled neural network, based on the phenomena of synchronous pulse bursts in the animal visual cortex, is different from traditional artificial neural networks. This paper first introduces a new approach for binary image thinning by using the pulse parallel transmission characteristic of PCNN. The thinning result obtains when pulses emitted by background meet. The criterion of pulse meeting and the criterion of thinning completion are proposed. The computer simulation results of applying the method to thin binary image are present. Comparisons of skeleton structure and execution time with results from other thinning methods are present too. The PCNN skeleton retains more information of original binary image, such as the size of a quadrate, than the result from Zhang and Suen method. The procedure is faster than Arcelli et al. thinning method when the image resolution is from 600 to1800 dpi. Combining with PCNN restoration algorithm (namely PCNN noise-reducing algorithm), the skeletons of the objects in a noisy binary image can be obtained with the accuracy. This paper also expands the application field of PCNN.

A quantitative structural study of fresh cement paste by image analysis: Part 2 : Measurements and their application

This paper is the second part of a study of cement paste structure by image analysis. Solid concentration and granulometric measurement have been carried out on the original binary image, then on the same image whose morphology had been modified. These results revealed anhydrous cement grains dispersed in the fresh paste. The effect of concentration and the state of the solid phase can influence this dispersion and we suggest that our method can be used to investigate both these phenomena.

Gray levels can improve the performance of binary image digitizers

The application of gray-scale digitizers to digitization of binary images of straight-edged planar shilhouettes is considered. A measure of digitization-induced ambiguity is introduced. It is shown that if the gray levels are not quantized and the spatial sampling resolution is sufficiently high, error-free reconstruction of the original binary image from the digitized image is possible. When the total bit-count for the representation of the digitized image is limited, i.e., sampling resolution and quantization accuracy are both finite, error-free reconstruction is usually impossible. In this case a bit allocation problem arises, and it is shown that the sensible bit allocation policy is to increase the quantization accuracy as much as possible once a “sufficient” spatial sampling resolution has been reached.

Fast line detection algorithm using multivalued image coding

A fast vision algorithm is proposed for detecting geometrical features such as line slopes of a polygonal object by using a multivalued image. Based on the fact that the multivalued image holds the local topological property of the original binary image, the algorithm extracts features by simple summation of weighted values of pixels in the multivalued image. Owing to the fact that the algorithm is easily implementable in hardware, its performance is experimentally shown to be very efficient in comparison with those of other methods.

機械製図図面の認識 (第1報)

This paper presents a method of bottom up processing for separating thick lines and hair lines and for recognizing straight lines, circles and arcs. It plays a fundamental and important role in the whole recognition method for mechanical drawings written on the basis of JIS drawing rules. This is the first report preceding other issues such as generating a recognition model, top down verify/modify processing and dimension matching for generation of a 2-D CAD model. First, all original binary image data are classified into two classes, thick and hair lines, with use of line width. Then, they are transformed into skeletons in both classes and after that each skeleton is divided into several segments. Further, based on the geometrical feature of the segments, those of thick lines are classified into straight lines, circles, arcs, etc., and they are transformed into structural representations. Some examples are finally shown.