Current Pixel Value Research Articles

Approximate bilateral filters for real-time and low-energy imaging applications on FPGAs

Bilateral filtering is an image processing technique commonly adopted as intermediate step of several computer vision tasks. Opposite to the conventional image filtering, which is based on convolving the input pixels with a static kernel, the bilateral filtering computes its weights on the fly according to the current pixel values and some tuning parameters. Such additional elaborations involve nonlinear weighted averaging operations, which make difficult the deployment of bilateral filtering within existing vision technologies based on real-time and low-energy hardware architectures. This paper presents a new approximation strategy that aims to improve the energy efficiency of circuits implementing the bilateral filtering function, while preserving their real-time performances and elaboration accuracy. In contrast to the state-of-the-art, the proposed technique allows the filtering action to be on the fly adapted to both the current pixel values and to the tuning parameters, thus avoiding any architectural modification or tables update. When hardware implemented within the Xilinx Zynq XC7Z020 FPGA device, a 5 × 5 filter based on the proposed method processes 237.6 Mega pixels per second and consumes just 0.92 nJ per pixel, thus improving the energy efficiency by up to 2.8 times over the competitors. The impact of the proposed approximation on three different imaging applications has been also evaluated. Experiments demonstrate reasonable accuracy penalties over the accurate counterparts.

WITHDRAWN: Strip lines and blotches noise removal using dynamic artifacts removal algorithm for HD images

Abstract The proposed method demonstrates the removal of different degradations that occur along with impulse noise in images. It consists of two steps; first, it detects the noise variance. Second, it filters all the degradation along with impulse noise. The current pixel value is replaced by a processed value in the filter’s output when the point of pixel is centered at that time. A prime benefit to this adaptive approach is that it not only wears down away edges or other small structures in the image but also replaces a number of autonomous algorithms required for removal of various degradations with impulse noise. The performance of the algorithm is calculated in terms of different metrics like MSE, PSNR and MSSIM.

Reversible Data Hiding in Encrypted Images Based on Block-Wise Multi-Predictor

Reversible data hiding of encrypted images (RDHEI) is a technology where after the original image is encrypted with a key, secret data can be embedded in the encrypted image without knowing the information of the original image. In recent years, RDHEI has attracted much attention due to its security and effectiveness, and many RDHEI methods have been proposed. However, the capacity of embedded data is still not high enough. In this paper, a high-capacity RDHEI algorithm based on block-wise multi-predictor, and improved Huffman coding is proposed. First, the original image is divided into non-overlapping blocks. To better exploit the spatial correlation of the original image, 16 prediction models are used to predict the current pixel value from the neighboring pixels. All the pixels in one block use the same prediction model. The improved Huffman encoding is used to compress the prediction error image and the optimal prediction model determined by the sum of the absolute prediction errors, which makes more room for data embedding. The experimental results show that our method is superior to some of the latest algorithms in terms of achieving a higher embedding capacity.

AN EFFICIENT CONFUSION-DIFFUSION STRUCTURE FOR IMAGE ENCRYPTION USING PLAIN IMAGE RELATED HENON MAP

Chaos-based image encryption has great significance as a branch of image security. So, a series of chaos-based cryptosystems protecting digital images are proposed in recent years. But, most of them have been broken as a result of poor encryption structure. This research paper suggests an effective image encryption structure to resist possible attacks. The proposed method employs plain image related Henon map (PIHM) for shuffling and diffusion processes in a connected way which is different from conventional chaotic based image encryption systems, since the initial conditions of diffusion process are established based on the initial conditions of shuffling process. The principle of confusion is achieved by shuffling the pixels over all the rows and columns. And the diffusion is ensured by using XOR operation of current shuffled pixel value with the previous value, and random pixel produced from PIHM map. The results of simulation and security analysis indicate that the proposed scheme has desirable encryption effects and is robust against different common attacks.

Texture image retrieval using hybrid directional Extrema pattern

In this paper, a new descriptor Hybrid Directional Extrema Pattern (HDEP) for the retrieval of texture images by integrating the concept of Weight Difference Directional Local Extrema Pattern (WD_DLEP) and Directional Local Extrema Pattern (DLEP) is proposed. The texture patterns are computed for four principle directions i.e., 0°, 45°, 90°, 135°. The proposed approach considers the difference between central pixel and corresponding neighboring pixels in the specified directions. This difference is used as weight in next stage. This weight is compared with a user-defined threshold to determine the value of strong bits in a feature vector. Experimental evaluation on three benchmark datasets (Brodatz, VisTex and Describable Textures Dataset) illustrates the better performance of proposed system with the other state-of-the-art techniques on two basic parameters i.e., average retrieval rate and time. The proposed approach is evaluated against various state-of-the-art texture image-retrieval systems based on local binary pattern, directional local extrema pattern, local tetra pattern, block-based local binary pattern, center-symmetric local binary pattern and wavelet. Significant improvement has been achieved in image retrieval performance due to assignment of weight in the pattern generation process. Further, the proposed approach is capable of differentiating different texture patterns more efficiently because it uses magnitude of pixel differences to determine the value of current pixel rather than the sign of pixel differences.

An efficient general data hiding scheme based on image interpolation

Information exchange through public networks has the risk of information disclosure, making data hiding an important technique to protect privacy. Image interpolation is also a common technique in image processing. Recent years, research on data hiding based on image interpolation have matured. In this paper, we propose an efficient data hiding scheme based on image interpolation that can be adopted into different image interpolation techniques. Different from other schemes, the converted decimal secret data is embedded into the cover image by adding or subtracting the cover pixels according to whether the current pixel value is lower or higher than the original pixel value and is embedded following the block complexity from high to low. Therefore, our proposed scheme keeps a relative balance between hiding capacity and image quality that can increase the stego-image quality and obtain a relatively higher image quality when the embedded bits are lower. The experimental results show that our proposed scheme has a relatively higher hiding capacity and can achieve almost a 1.6 dB increase on average compared with other schemes. The visual effect is also superior under the different image interpolation techniques.

Ameliorated mean adaptive median filtering algorithm

Aiming at the problem that the quality of infrared image decreases due to the large amount of random noise in the process of collection and transmission of infrared image of electrical equipment, and the accuracy of automatic detection of electrical equipment decreases, based on the traditional adaptive median filter algorithm, the adaptive median filter is analyzed, which can filter only the salt and pepper noise below 25%. An improved mean adaptive median filtering algorithm is proposed to overcome the shortcomings of wave effect. Firstly, the filtering window is selected according to the decision setting condition, and then it is judged whether the K-mean value near the center point is a noise point, and if so, the window is increased, otherwise the average value is output. Finally, it is judged whether the value of the current pixel point is noise, and if so, the average value is output, otherwise, the current pixel value is output. The experimental results show that the algorithm can effectively filter salt and pepper noise and Gauss noise, while maintaining the image sharpness, and has good filtering performance on PSNR and MSE indicators.

Application of Improved Neural Network Algorithm in Image Denoising and Edge Detection

The knowledge involved in digital image processing is very wide, and there are many kinds of methods. Traditional image processing technology is mainly focused on the acquisition, transformation, enhancement, restoration, compression encoding, segmentation, edge extraction and so on. With the emergence of new tools and new methods, the image processing technology has been updated and developed. In this paper, an effective method for edge detection and image de-noising is proposed. In this article, the impulse noise detector is composed of a BP neural network (BPNN) and a decision switch. BPNN requires four input values, which are the current pixel value, grey median value, energy value, and contrast. To take these four values as the input values, the impulse noise detector can show good performance. The output of the BPNN is transferred to the decision switch, and the output value is converted to 0 or 1, which is used to distinguish whether the pixels are polluted. At this point, we introduce an additional impulse term and establish the improved BPNN model. The additional impulse term can effectively speed up the convergence of the network, avoid the emergence of the local minimum problem, and ensure the stability of the training process. In this way, the IBPNN filter of this paper only uses the information of the non polluted pixels to filter the noise pixels, which avoids the secondary pollution, and obtains a better performance. This algorithm has high PSNR value and strong detail information and edge preserving ability. Finally, the improved BPNN algorithm is applied to the image edge detection, and we use the improved neural network model to detect the edge of the image. Because the method can be used to include the prior knowledge, the IBPNN method is better than the traditional method in image edge detection.

The Research of Feature Extraction Method of Liver Pathological Image Based on Multispatial Mapping and Statistical Properties.

We propose a new feature extraction method of liver pathological image based on multispatial mapping and statistical properties. For liver pathological images of Hematein Eosin staining, the image of R and B channels can reflect the sensitivity of liver pathological images better, while the entropy space and Local Binary Pattern (LBP) space can reflect the texture features of the image better. To obtain the more comprehensive information, we map liver pathological images to the entropy space, LBP space, R space, and B space. The traditional Higher Order Local Autocorrelation Coefficients (HLAC) cannot reflect the overall information of the image, so we propose an average correction HLAC feature. We calculate the statistical properties and the average gray value of pathological images and then update the current pixel value as the absolute value of the difference between the current pixel gray value and the average gray value, which can be more sensitive to the gray value changes of pathological images. Lastly the HLAC template is used to calculate the features of the updated image. The experiment results show that the improved features of the multispatial mapping have the better classification performance for the liver cancer.

A novel reversible image data hiding scheme based on pixel value ordering and dynamic pixel block partition

Recently, various efficient reversible data-hiding schemes based on pixel value ordering have been proposed for embedding messages into high-fidelity images. In these schemes, after dividing the cover image into equal-sized blocks, the pixels within a given block are ordered according to their values, and data embedding is achieved by modifying the maximum and minimum values of each block. For a given embedding capacity, the optimal block size is exhaustively searched so that the embedding distortion is minimized. These pixel value ordering-based schemes perform fairly well, especially for low embedding capacity. However, to obtain a larger embedding capacity, a smaller block size should be used, which usually leads to a dramatic quality degradation of the marked image. In this paper, to address this drawback and to enhance the performance of pixel value ordering-based embedding further, a novel reversible data hiding method is proposed. Instead of using equal-sized blocks, a dynamic blocking strategy is used to divide the cover image adaptively into various-sized blocks. Specifically, flat image areas are preferentially divided into smaller blocks to retain high embedding capacity, whereas rough areas are divided into larger blocks to avoid decreasing peak signal-to-noise ratio. As a result, the proposed scheme can provide a larger embedding capacity than current pixel value ordering-based schemes while keeping distortion low. The superiority of the proposed method is also experimentally verified.

OBJECT DETECTION USING REAL TIME ALGORITHM WITH FACE RECOGNITION

A real and accurate object detection and Recognition technique is used to detect the real objects in the human face such as eye, nose etc…The face recognition is applicable to detect only the exact human face. But it does nose recognize the human face if there is any damage found in the face. To overcome these problem implement the Bayesian algorithm is used to detect the objects on human face like eye, nose etc... By using these algorithm the correct authorized user can be identified by comparing the current pixel values of the real object with the pixel values already stored in the database This system use C4 algorithm along with the MSRF (Multi Scale Random Field) because these methods are used to calculate only the neigbhouring pixel value by providing the fast rejection on the background image .Grayscale conversion from an RGB value results in the easy calculation of the pixel value. Pixel by pixel calculation is done and provides the secure authorization for the user against the hackers because the pixel value of an particular authorized person is entirely differ from the hackers, which can be get detected by SMAP (Sequential MAP)while handling transaction in an ATM Machine.

Partition-based fuzzy median filter based on adaptive resonance theory

This paper presents a novel partition-based fuzzy median filter for noise removal from corrupted digital images. The proposed filter is obtained as the weighted sum of the current pixel value and the output of the median filter, where the weight is set by using fuzzy rules concerning the state of the input signal sequence to indicate to what extent the pixel is considered to be noise. Based on the adaptive resonance theory, the authors developed a neural network model and created a new weight function where the neural network model is employed to partition the observation vector. In this framework, each observation vector is mapped to one of the M blocks that form the observation vector space. The least mean square (LMS) algorithm is applied to obtain the optimal weight for each block. Experiment results have confirmed the high performance of the proposed filter in efficiently removing impulsive noise and Gaussian noise.

Spatial Information Based on the Multi-Objective Programming of Fuzzy Kernel Clustering Image Segmentation Algorithm

Fuzzy kernel clustering algorithm is a combination of unsupervised clustering and fuzzy set of the concept of image segmentation techniques, But the algorithm is sensitive to initial value, to a large extent dependent on the initial clustering center of choice, and easy to converge to local minimum values, when used in image segmentation, membership of the calculation only consider the current pixel values in the image, and did not consider the relationship between neighborhood pixels, and so on segmentation contains noise image is not ideal. This paper puts forward an improved fuzzy kernel clustering image segmentation algorithm, the multi-objective problem, change the single objective problem to increase the secondary goals concerning membership functions, Then add the constraint information space; Finally, using spatial neighborhood pixels corrected membership degree of the current pixel. The experimental results show that the algorithm effectively avoids the algorithm converges to local extremism and the stagnation of the iterative process will appear problem, significantly lower iterative times, and has good robustness and adaptability.

Watermark with DSA signature using predictive coding

This paper presents a predictor-based watermark scheme that embeds secret bit streams and a DSA signature into an image. For the copyrighting of digital media, a DSA signature is appropriate as a watermarking technique. To improve security, we apply an Arnold transform (AT) to secret messages. We propose new predictors, Left-Top, which the predict current pixel values using neighboring pixel values. Our proposed scheme conceals secret messages by using the difference between current pixel values and predictive pixel values. Experimental results show that our method has low complexity and achieves a higher embedding performance with good perceptual quality compared to the earlier arts. Experimental results verified our proposed watermark method in multimedia communications.

Data hiding based on two-stage referencing for two-colour images

Data hiding in two-colour images is difficult since 1 pixel requires only 1 bit representation and it is easy to detect for pixel distortion. In this paper, we describe a new data hiding method for two-colour images by two-stage referencing. The cover image is partitioned into n×n non-overlapping sub-blocks, and we calculate difference values by two stages for all pixels to find the suitable replacement pixel. The two-stage referencing is to obtain difference values that are different with the current pixel value for neighbouring pixels with n×n and (n+2)×(n+2). These two difference values are used to embed a secret bit on the sub-block. The experimental results show that the proposed method achieves a good visual quality for the stego-image.

Automated Surveillance System and Data Communication

Automated video surveillance systems plays vital role in the business areas , military system and etc . Many research areas in video surveillance system is mainly focuses on algorithms to evaluate background subtraction system and alert to the system, to repeatedly detect major actions For example, our proposed technique stores, for each pixel, a set of values taken in the past at the same location or in the neighborhood. It then compares this set to the current pixel value in order to determine whether that pixel belongs to the background, and adapts the model by choosing randomly which values to substitute from the background model. This approach differs from those based upon the classical belief that the oldest values should be replaced first. In background comparison if the difference in the values reaches the threshold value then user can get alert message. So whenever user getting SMS from server system and the background image is updated whenever the system is detecting a motion.

An improved human visual system based reversible data hiding method using adaptive histogram modification

Jung et al., IEEE Signal Processing Letters, 18, 2, 95, 2011 proposed a reversible data hiding method considering the human visual system (HVS). They employed the mean of visited neighboring pixels to predict the current pixel value, and estimated the just noticeable difference (JND) of the current pixel. Message bits are then embedded by adjusting the embedding level according to the calculated JND. Jung et al.'s method achieved excellent image quality. However, the embedding algorithm they used may result in over modification of pixel values and a large location map, which may deteriorate the image quality and decrease the pure payload. The proposed method exploits the nearest neighboring pixels to predict the visited pixel value and to estimate the corresponding JND. The cover pixels are preprocessed adaptively to reduce the size of the location map. We also employ an embedding level selection mechanism to prevent near-saturated pixels from being over modified. Experimental results show that the image quality of the proposed method is higher than that of Jung et al.'s method, and the payload can also be increased due to the reduction of the location map.

A robust watermark authentication technique based on Weber’s descriptor

One of the major challenges in the field of digital image watermarking is to authenticate the presence of watermark in the watermarked image even after it has been transformed intentionally or unintentionally. Transformation can be geometric-like rotation, scaling, and translation of image or may be due to any signal processing attack like noise corruption, compression, and cropping of image. There may also be some photometric changes, for example change in the brightness of watermarked image during transmission, due to which it becomes difficult to validate whether received image is watermarked or not. Illumination invariance property of Weber’s descriptor has engrossed to use it in the proposed watermark authentication technique. Weber’s descriptor is a descriptor based on two parameters of a pixel, differential excitation and orientation. These parameters are computed using the relative intensity value of neighbor pixels and current pixel. This descriptor remains the same even after intensity changes due to the contribution of all neighbor pixel’s intensity in its computation. It is also known to be robust to scaling and rotation. Experimental results show that the proposed watermarking technique is able to authenticate the presence of watermark in the watermarked image even when it is distorted due to geometric and photometric attacks. In addition to this, it is found to be robust against noise, cropping, and compression attacks.

An Efficient Neighbourhood Pixel Filtering Algorithm for Wavelet-Based Image Denoising

Image denoising using the wavelet transform has been attracting much attention. Image corrupted by a noise is a classical problem in the field of signal or image processing. À Trous algorithm is introduced to overcome the problem of translation variant mechanism, existing on using discrete wavelet transform (DWT). This algorithm up samples low-pass filter by inserting zeros between the filter coefficients at each level and accordingly the low-pass and high-pass filter coefficients are modified. The efficiency of wavelet images by using this algorithm is low because the detail preservations of images at different scales are not uniform; also random noise rapidly attenuates with increasing scales. Due to this the contrast of the resulting image is weaker. In order to improve the clarity of the image an algorithm called a neighbourhood pixel filtering algorithm (NPFA) is added along with the existing à trous algorithm. In the proposed algorithm, find neighbourhood pixel difference (NPD ) by subtracting the neighbourhood pixel values from its current noisy pixel value. Also, calculate weight of each pixel which depends on this NPD. A filtered value is assigned for each current pixel in order to approximate the original pixel value of that pixel. This filtered value is generated by minimizing NPD and weighted mean square error (WMSE) using method of least square. A reduction in noise pixel is observed on replacing the optimal weight namely NPFA filter solution with the noisy value of the current pixel. Due to this NPFA filter gain the effect of both high-pass and low-pass filter. This filter behaves like a low-pass filter in smooth region by decreasing noise variance effectively and giving similar weights to all its neighbourhood pixels. This in turn cuts off only high frequency noise signal instead of all noisy signals. The resultant image thus obtained is observed to have much less blurring effect compared to the other wavelet method.

ViBe: A Universal Background Subtraction Algorithm for Video Sequences

This paper presents a technique for motion detection that incorporates several innovative mechanisms. For example, our proposed technique stores, for each pixel, a set of values taken in the past at the same location or in the neighborhood. It then compares this set to the current pixel value in order to determine whether that pixel belongs to the background, and adapts the model by choosing randomly which values to substitute from the background model. This approach differs from those based upon the classical belief that the oldest values should be replaced first. Finally, when the pixel is found to be part of the background, its value is propagated into the background model of a neighboring pixel. We describe our method in full details (including pseudo-code and the parameter values used) and compare it to other background subtraction techniques. Efficiency figures show that our method outperforms recent and proven state-of-the-art methods in terms of both computation speed and detection rate. We also analyze the performance of a downscaled version of our algorithm to the absolute minimum of one comparison and one byte of memory per pixel. It appears that even such a simplified version of our algorithm performs better than mainstream techniques.