Adnan Mustafa | Kuwait University (original) (raw)

Papers by Adnan Mustafa

The problem of selecting a best representative image from a group of similar images is an importa... more The problem of selecting a best representative image from a group of similar images is an important problem as it can expedite the task of image search and image matching. We solve this problem by first measuring the similarity between every pair of image in the set by a suitable similarity measure, and then transforming the problem to similarity space and finding the corresponding locations of the images in the similarity space. Finally, the image located closest to the center of the preoccupied similarity space is selected as the best representative image. The difficulty in such a problem arises in attempting to find the locations of N images in the similarity space, since this leads to a set of N(N–1) non-linear simultaneous algebraic equations with N unknowns. We solve such a problem by forcing the solution to be in . We present a closed-form solution for the cases when N = 3, 4 and 5. We give examples of finding the best representative images for two sets as an application of ...

In this paper we present the Probabilistic Matching Model for Binary Images (PMMBI), a model for ... more In this paper we present the Probabilistic Matching Model for Binary Images (PMMBI), a model for the quick detection of dissimilar binary images based on random point mappings. The model predicts the probability of detecting dissimilarity between any pair of binary images based on the amount of similarity and number of random pixel mappings between them. Based on the model, we show that by performing a limited number of random pixel mappings between binary images, dissimilarity detection can be performed quickly. Furthermore, the model is image size invariant; the size of the image has absolutely no effect on the dissimilarity detection quickness. We give examples with real images to show the accuracy of the model.

WSEAS TRANSACTIONS ON MATHEMATICS, 2021

Applications of Digital Image Processing XXXIX, 2016

A Probabilistic Matching Model for Binary Images (PMMBI) is presented that predicts the probabili... more A Probabilistic Matching Model for Binary Images (PMMBI) is presented that predicts the probability of matching binary images with any level of similarity. The model relates the number of mappings, the amount of similarity between the images and the detection confidence. We show the advantage of using a probabilistic approach to matching in similarity space as opposed to a linear search in size space. With PMMBI a complete model is available to predict the quick detection of dissimilar binary images. Furthermore, the similarity between the images can be measured to a good degree if the images are highly similar. PMMBI shows that only a few pixels need to be compared to detect dissimilarity between images, as low as two pixels in some cases. PMMBI is image size invariant; images of any size can be matched at the same quick speed. Near-duplicate images can also be detected without much difficulty. We present tests on real images that show the prediction accuracy of the model.

Indian Journal of Science and Technology, 2016

Given two binary images, how can we determine if the images are different? The most common techni... more Given two binary images, how can we determine if the images are different? The most common technique used in the image processing arena is to calculate the correlation between the two images or simply to subtract the two images. Both of these methods require some type of operation to be applied to the whole image. This implies that as the image size increases, more time will be required. In this paper, we show that only a limited number of points need to be checked to arrive at a required confidence level that the images are the same or different. In fact, for completely different binary images only 8 points need to be checked to arrive at a 99% confidence level, 11 points need to be checked to arrive at a 99.9% confidence level and only 15 points need to be checked to arrive at a 99.99% confidence level that the images are the same or different. As a result, this method is magnitudes faster than traditional methods.

WSEAS Transactions on Systems and Control archive, 2017

Many probabilistic models have been developed for numerous problems in robot and computer vision ... more Many probabilistic models have been developed for numerous problems in robot and computer vision such as for image segmentation, road extraction, and object tracking. In this paper we present a probabilistic model for the random pixel mapping of binary images. The model predicts the probability of detecting dissimilarity between dissimilar binary images as a function of the number of random mappings and the amount of similarity. The model shows that detecting dissimilarity can be accomplished quickly by random pixel mapping, without the need to process the entire image. Test results on real images are presented that show the accuracy of the model.

IET Image Processing

Here, the author presents the gamma binary distance, an exceptional distance for measuring simila... more Here, the author presents the gamma binary distance, an exceptional distance for measuring similarity between binary images. The gamma distance is a probabilistic pixel mapping measure that is a modification of the Hamming distance. Employing a probabilistic approach to image matching enables gamma to measure similarity more accurately than employing traditional binary distances. The author shows the advantage of employing the gamma distance for similarity measurement by comparing it to three of the most popular similarity distances used for binary image matching: correlation, sum of the absolute difference method, and mutual information. Results of extensive testing conducted on a large database are presented where the superiority of the gamma distance over other similarity distances is shown.

SPIE Proceedings

In previous work, a probabilistic image matching model for binary images was developed that predi... more In previous work, a probabilistic image matching model for binary images was developed that predicts the number of mappings required to detect dissimilarity between any pair of binary images based on the amount of similarity between them. The model showed that dissimilarity can be detected quickly by randomly comparing corresponding points between two binary images. In this paper, we improve on this quickness for images that have dissimilarity concentrated near their centers. We apply smart mapping schemes to different image sets and analyze the results to show the effectiveness of this mapping scheme for images that have dissimilarity concentrated near their center. We compare three different smart mapping schemes with three different mapping densities to find the best mapping / best density performance.

World Academy of Science, Engineering and Technology, International Journal of Computer, Electrical, Automation, Control and Information Engineering, 2018

In this paper we present a quick technique to measure the similarity between binary images. The t... more In this paper we present a quick technique to measure the similarity between binary images. The technique is based on a probabilistic mapping approach and is fast because only a minute percentage of the image pixels need to be compared to measure the similarity, and not the whole image. We exploit the power of the Probabilistic Matching Model for Binary Images (PMMBI) to arrive at an estimate of the similarity. We show that the estimate is a good approximation of the actual value, and the quality of the estimate can be improved further with increased image mappings. Furthermore, the technique is image size invariant; the similarity between big images can be measured as fast as that for small images. Examples of trials conducted on real images are presented. Keywords—Big images, binary images, similarity, matching.

WSEAS TRANSACTIONS on MATHEMATICS, 2021

The task of data matching arises frequently in many aspects of science. It can become a time cons... more The task of data matching arises frequently in many aspects of science. It can become a time consuming process when the data is being matched to a huge database consisting of thousands of possible candidates, and the goal is to find the best match. It can be even more time consuming if the data are big (> 100 MB). One approach to reducing the time complexity of the matching process is to reduce the search space by introducing a pre-matching stage, where very dissimilar data are quickly removed. In this paper we focus our attention to matching big binary data. In this paper we present two probabilistic models for the quick dissimilarity detection of big binary data: the Probabilistic Model for Quick Dissimilarity Detection of Binary vectors (PMQDD) and the Inverse-equality Probabilistic Model for Quick Dissimilarity Detection of Binary vectors (IPMQDD). Dissimilarity detection between binary vectors can be accomplished quickly by random element mapping. The detection technique is not a function of data size and hence dissimilarity detection is performed quickly. We treat binary data as binary vectors, and hence any binary data of any size and dimension is treated as a binary vector. PMQDD is based on a binary similarity distance that does not recognize data and its exact inverse as containing the same pattern and hence considers them to be different. However, in some applications a specific data and its inverse, are regarded as the same pattern, and thus should be identified as being the same; IPMQDD is able to identify such cases, as it is based on a similarity distance that does not distinguish between data and its inverse instance as being dissimilar. We present a comparative analysis between PMQDD and IPMQDD, as well as their similarity distances. We present an application of the models to a set of object models, that show the effectiveness and power of these models..

International Conference on Signal, Image Processing and Applications, 2019

Image matching is a fundamental problem that arises frequently in many aspects of robot and compu... more Image matching is a fundamental problem that arises frequently in many aspects of robot and computer vision. It can become a time consuming process when matching images to a database consisting of hundreds of images-if exact matching is required. It can be even more time consuming if the images are big. One approach to reducing the time complexity of the matching process is to reduce the search space by introducing a pre-matching stage, where dissimilar images are quickly removed. The Probabilistic Matching Model for Binary Images (PMMBI) showed that dissimilarity detection between binary images can be accomplished quickly by random pixel mapping and is size invariant. The model is based on the gamma binary similarity distance that recognizes an image and its inverse as containing the same scene and hence considers them to be the same image. However, in many applications, an image and its inverse are not treated as being the same but rather different. In this paper, we present a comparative analysis of dissimilarity detection between PMMBI based on the gamma binary similarity distance and a modified PMMBI model based on a modified gamma binary similarity distance that does distinguish between an image and its inverse as being dissimilar.

International Journal of Engineering Research and Technology, 2019

In this paper we present a fast size-invariant method for binary image matching. The method, call... more In this paper we present a fast size-invariant method for binary image matching. The method, called Dissimilar Detection via Mapping (DDM), is based on probabilistic matching models that can quickly detect dissimilarity between binary images regardless of image size. Dissimilarity detection is performed quickly by comparing only a few points (pixels) between the images. As a result, image matching can be performed fairly quickly. A complete detailed analysis of DDM and the mathematical proof of its superiority over other methods as images become big is given. We compare DDM to three of the most popular matching methods employed in the image processing arena: image correlation, sum of the absolute difference and mutual information. We show that DDM is magnitudes faster than these methods-if the images are not small. Furthermore, we show how DDM can be used as a pre-processor for other matching methods to speed up their matching speed. In particular, we use DDM with image correlation to enhance the latter's performance. Test results are presented for real images varying in size from 16 kilo-pixels to 10 mega-pixels to show the quickness of the method and its size-invariance.

Proc. SPIE 9971, Applications of Digital Image Processing XXXIX, 2016

The problem of selecting a best representative image from a group of similar images is an importa... more The problem of selecting a best representative image from a group of similar images is an important problem as it can expedite the task of image search and image matching. We solve this problem by first measuring the similarity between every pair of image in the set by a suitable similarity measure, and then transforming the problem to similarity space and finding the corresponding locations of the images in the similarity space. Finally, the image located closest to the center of the preoccupied similarity space is selected as the best representative image. The difficulty in such a problem arises in attempting to find the locations of N images in the similarity space, since this leads to a set of N(N–1) non-linear simultaneous algebraic equations with N^2 unknowns. We solve such a problem by forcing the solution to be in R^(N–1). We present a closed-form solution for the cases when N = 3, 4 and 5. We give examples of finding the best representative images for two sets as an application of the method.

Given two sets of binary data, how can we determine if the data are dissimilar? The simplest tech... more Given two sets of binary data, how can we determine if the data are dissimilar? The simplest technique is to simply subtract the two sets or to calculate the correlation between them. Both of these methods –as well as other methods– require some type of similarity operation to be applied to all points of the data. This implies that as the data becomes big, more processing
time is required. In this paper, we present a novel approach to matching using a probabilistic model that requires a few number of points –and not all points – to be compared between two data sets to detect dissimilarity. Furthermore, the model is size invariant; big data can be matched just as quickly as matching small data. The similarity between the data can also be measured to a good degree by repeating the matching process several times.

WSEAS TRANSACTIONS ON MATHEMATICS, 2021

Applications of Digital Image Processing XXXIX, 2016

Indian Journal of Science and Technology, 2016

WSEAS Transactions on Systems and Control archive, 2017

IET Image Processing

SPIE Proceedings

World Academy of Science, Engineering and Technology, International Journal of Computer, Electrical, Automation, Control and Information Engineering, 2018

WSEAS TRANSACTIONS on MATHEMATICS, 2021

International Conference on Signal, Image Processing and Applications, 2019

International Journal of Engineering Research and Technology, 2019

Proc. SPIE 9971, Applications of Digital Image Processing XXXIX, 2016

The problem of selecting a best representative image from a group of similar images is an importa... more The problem of selecting a best representative image from a group of similar images is an important problem as it can expedite the task of image search and image matching. We solve this problem by first measuring the similarity between every pair of image in the set by a suitable similarity measure, and then transforming the problem to similarity space and finding the corresponding locations of the images in the similarity space. Finally, the image located closest to the center of the preoccupied similarity space is selected as the best representative image. The difficulty in such a problem arises in attempting to find the locations of N images in the similarity space, since this leads to a set of N(N–1) non-linear simultaneous algebraic equations with N^2 unknowns. We solve such a problem by forcing the solution to be in R^(N–1). We present a closed-form solution for the cases when N = 3, 4 and 5. We give examples of finding the best representative images for two sets as an application of the method.