Khalil Alsaif - Academia.edu (original) (raw)

Papers by Khalil Alsaif

Nucleation and Atmospheric Aerosols, 2023

Object tracking is regarded as one of the significant topics in the scope of computer vision, whi... more Object tracking is regarded as one of the significant topics in the scope of computer vision, which led to rapid development in the practical field through enhancing the reliable tools available in such a field.Recently, the appearance of Artificial Neural Networks (ANNs) resulted in new methods to the identification and recognition of objects. In this research a noble method is suggested for tracking objects inside video files, the method was used to detect cars on the streets through extracting the features of these cars by using You Only Look Ones (YOLO v3) Our method was able to process the extracted features of the cars at detection phase. Also, it exploited the extracted feature in an effective way in order to obtain precise detections. The findings show that our proposed method performs better than the other methods available in the field, as it is able to produce better predictions by using less computation which resulted in reducing the time that such process usually takes. The evaluation results show that our method was able to process an average of 207.6 frames per second to track objects with 67.6% Multi-Object Tracking Accuracy (MOTA) and 89.1% Multi-Object Tracking Precision (MOTP).

2ND INTERNATIONAL CONFERENCE ON MATHEMATICAL TECHNIQUES AND APPLICATIONS: ICMTA2021

Object tracking is regarded as one of the significant topics in the scope of computer vision, whi... more Object tracking is regarded as one of the significant topics in the scope of computer vision, which led to rapid development in the practical field through enhancing the reliable tools available in such a field.Recently, the appearance of Artificial Neural Networks (ANNs) resulted in new methods to the identification and recognition of objects. In this research a noble method is suggested for tracking objects inside video files, the method was used to detect cars on the streets through extracting the features of these cars by using You Only Look Ones (YOLO v3) Our method was able to process the extracted features of the cars at detection phase. Also, it exploited the extracted feature in an effective way in order to obtain precise detections. The findings show that our proposed method performs better than the other methods available in the field, as it is able to produce better predictions by using less computation which resulted in reducing the time that such process usually takes. The evaluation results show that our method was able to process an average of 207.6 frames per second to track objects with 67.6% Multi-Object Tracking Accuracy (MOTA) and 89.1% Multi-Object Tracking Precision (MOTP).

Journal of Al-Qadisiyah for Computer Science and Mathematics

Reliance on computer vision systems in the sports field is one of the very important topics, whic... more Reliance on computer vision systems in the sports field is one of the very important topics, which are of high importance, especially in the arbitration process or evaluating the accuracy of the player’s performance of the movement. It is better to rely on computer vision systems that are more accurate in the arbitration process. In this article, a method was presented to distinguish one of the important movements of the gymnastics player, by relying on deep learning techniques. The dataset was built based on high-quality video clips found on YouTube for tournaments held from the period 2018-2022, due to the absence of The dataset available. This data was divided into three sections: 70% for training, 10% for validation, and 20% for testing. Two models of the convolutional neural network yolov7 and yolov5 were trained, and the results obtained after testing the results of the models show that the seventh version was the best , Recall, Precision and Mean Average Precision criteria we...

The methods of detecting objects and tracking their movements are among the methods that are reli... more The methods of detecting objects and tracking their movements are among the methods that are relied upon in many fields, whether medical or industrial, and others. One of these areas that will rely on deep learning methods in discovering and distinguishing the player's movements is the sports field and is very useful in games in which the player's degree depends on the accuracy of the performance of the movement, such as the gymnastics game, where it was applied to the static ring gymnastics game, where the distinction of movements was discovered the stability in this game is based on a convolutional neural network. Models. The neural network was trained on five of the most important stability movements in this game after creating the data set based on a set of videos of tournaments held in the period from 2016-2022, where an average of 1500 images were obtained for each stability movement, which was divided into 80% for training and 20 % for testing, after training the convolutional neural network model, it was applied to a group of video clips for different tournaments. Many criteria were adopted to measure the efficiency of the model after training and practical application, which showed the efficiency of the proposed system.

Journal of Al-Qadisiyah for Computer Science and Mathematics

Due to progress in the field of deep learning in order to find and track objects through the use ... more Due to progress in the field of deep learning in order to find and track objects through the use of computer vision in the service of large segments of the population, as it was adopted in the field of serving people b with special needs for the sake of dialogue and implementation of many requests in this research, a series of commands for use by people with special needs with speech problems or paralysis, where the ability to use the eye blink is very useful for social communication, were developed. In this research, the orders needed by the target people with special needs were studied, and (11) commands were identified that can be increased according to the intended sample. A table of commands was built depending on the length of the eye blink period. By creating a modified CNN: Convolutional Neural Network structure and training it on 2 different database, deep learning was used to identify and determine if the eye is closed or open (mrlEye2018 and Closed Eye in the Wlid:CEW).It...

Wasit Journal of Pure sciences

The methods of detecting objects and tracking their movements are among the methods that are reli... more The methods of detecting objects and tracking their movements are among the methods that are relied upon in many fields, whether medical or industrial, and others. One of these areas that will rely on deep learning meth-ods in discovering and distinguishing the player's movements is the sports field and is very useful in games in which the player's degree depends on the accura-cy of the performance of the movement, such as the gymnastics game, where it was applied to the static ring gymnastics game, where the distinction of move-ments was discovered the stability in this game is based on a convolutional neu-ral network. Models. The neural network was trained on five of the most im-portant stability movements in this game after creating the data set based on a set of videos of tournaments held in the period from 2016-2022, where an aver-age of 1500 images were obtained for each stability movement, which was di-vided into 80% for training and 20 % for testing, after training t...

Tikrit Journal of Pure Science

Image transformation provide deep meaning about images feature, so many type of image transformat... more Image transformation provide deep meaning about images feature, so many type of image transformation are appear in the last decade years, one of them is curvelet transformation which improve the image processing techniques specially in field of feature extraction. Personal authentication adopt biometric information to be one of the major coefficients in this field. Palm print one of the main approaches for personal identification. So studying the moments extracted from coefficients of curvelet transform of palm print image adopted in order to get high efficient system for personalization systems. Two major phases are constructed in this research to adopt the moments of low frequency coefficient of the curvelet for personal identification. In the first phase a database was built for 50 persons by acquisition nine images for both hands (9 for left hand and 9for right hand). images are acquired and then processed to extract ROI (region of interest) by looking for the palm centroid then...

Journal of University of Anbar for Pure Science

The data hiding is one of the most important subject in field of computer science, so a lot of te... more The data hiding is one of the most important subject in field of computer science, so a lot of technique was developed and modified to satisfy the optimum lend of hiding. In this research the contourlet transformation coefficients were studied to decide which of them are suitable to embed data on it a lot of parameters of the contourlet coefficients can be discussed one of them is the coefficient energy.The research covered most of the suggested events which could be met during the embedding state, one of them the size of the cover in addition to the size of the information were studied. Applying the suggested idea on different type of image with different size (cover image and the message image) shows that the coefficients with low level of energy are suitable to embedded the information, and the retrieved cover and message are so closed to the original one. Key word-image processing; data security; watermarking; contourlet.

JOURNAL OF EDUCATION AND SCIENCE

In this research, the important features of recognizing speakers' identity by extracting features... more In this research, the important features of recognizing speakers' identity by extracting features of speakers voices are used. Voice is considered one of the vital factors adopted algorithm on samples of speakers' voices has been implemented by recording the voice, saving it on a file of the type wav. The voice is treated by using Hamming function to reduce error ratio. Feature of the voice samples have been extracted by taking the spectrum value of the sound signal of all the speakers. Two methods have been used (Ecledian distance and correlation factor) for comparing core of the convolution of the sample of speaker test (a training sample which is being recorded another time) and all the speakers samples who have already been saved on the data base. Applying the algorithm on different voices, ratio of error was very small while matching has increased by the increasing number of speaker. The algorithm has been implemented using mathlab.

IRAOI JOURNAL OF STATISTICAL SCIENCES, Jun 28, 2011

In this research the covariance matrix which was used in so many fields, its eigen values adopted... more In this research the covariance matrix which was used in so many fields, its eigen values adopted to be the main parameters for Latin printed character recognition. The idea was divided in two stages. The first stage is to generate the covariance matrix then, evaluate its eigen values to build main table for the whole latin characters in addition to the numeric values. The second stage is the recognition stage, which achieved to any character that was entered. The applied examples do not register any negative result. So it can be strongly recommended for printed character recognition.

IRAOI JOURNAL OF STATISTICAL SCIENCES, 2013

The 6 th Scientific Conference of the College of Computer Sciences & Mathematics pp [165-180] ] 1... more The 6 th Scientific Conference of the College of Computer Sciences & Mathematics pp [165-180] ] 165 [

In this research it has been compared between two methods used to recognize Arabic character, the... more In this research it has been compared between two methods used to recognize Arabic character, the first based on histogram technique to extract the characteristic for the images of Arabic characters, and used with Kohonen neural network to recognize the Arabic characters. The second method was reliance on the seven invariant moment that extracted from Arabic characters the adopted by back propagation neural network for Arabic character recognition. In the first method, the Kohonen neural network has been trained on vertical histogram values for a set of letters(ث، ح ، د، ر، س، ض، ط، غ، ق، ن), and also it has been tested on a set of noisy letters images, and the results were achieved true, and compatible with the previous one. Also, kohonen network has been trained using horizontal histogram for the same group of characters, and the results had lower closeness from the results that obtained with vertical histogram. The results were achieved true 66%. Again, several type of the backpropagation neural network has been trained in seven invariant moment of the Arabic characters(ا، ب ، ح ، ع ، س ، ظ ، ر ، ذ ، ف، ك) and it has been used several types of the training functions that common used with backpropagation neural network. The results were achieved true 75% . So, it is discern previously that depending on vertical histogram values with kohonen neural network gives higher closeness from horizontal histogram vales with kohonen neural network and seven invariant moment with back propagation neural network. In addition, when depending on the two histograms together, it was gives bad results with other neural network as well as kohonen neural network.

AL-Rafidain Journal of Computer Sciences and Mathematics, 2021

With the rapid growth of computer usage to extract the required knowledge from a huge amount of i... more With the rapid growth of computer usage to extract the required knowledge from a huge amount of information, such as a video file, significant attention has been brought towards multi-object detection and tracking. Artificial Neural Networks (ANNs) have shown outstanding performance in multi-object detection, especially the Faster R-CNN network. In this study, a new method is proposed for multi-object tracking based on descriptors generated by a neural network that is embedded in the Faster R-CNN. This embedding allows the proposed method to directly output a descriptor for each object detected by the Faster R-CNN, based on the features detected by the Faster R-CNN to detect the object. The use of these features allows the proposed method to output accurate values rapidly, as these features are already computed for the detection and have been able to provide outstanding performance in the detection stage. The descriptors that are collected from the proposed method are then clustered into a number of clusters equal to the number of objects detected in the first frame of the video. Then, for further frames, the number of clusters is increased until the distance between the centroid of the newly created cluster and the nearest centroid is less than the average distance among the centroids. Newly added clusters are considered for new objects, whereas older ones are kept in case the object reappears in the video. The proposed method is evaluated using the UA-DETRAC (University at Albany Detection and Tracking) dataset and has been able to achieve 64.8% MOTA and 83.6% MOTP, with a processing speed of 127.3 frames per second.

AL-Rafidain Journal of Computer Sciences and Mathematics, 2021

The interest in coding was very high because it is widely relied on in the security of correspond... more The interest in coding was very high because it is widely relied on in the security of correspondence and in the security of information in addition to the need to rely on it in the storage of data because it leads to a pressure in the volume of information when storing it. In this research, image transformation was used to encode gray or color images by adopting parameters elected from contourlet transformations for image. The color images are acquired into the algorithm, to be converted into three slices (the main colors of the image), to be disassembled into their coefficients through contourlet transformations and then some high frequencies in addition to the low frequency are elected in order to reconstruct the image again. The election of low frequencies with a small portion of the high frequencies has led to bury some unnecessary information from the image components. The performance efficiency of the proposed method was measured by MSE and PSNR criteria to see the extent of the discrepancy between the original image and the recovered image when adopting different degrees of disassembly level, in addition, the extent to which the image type affects the performance efficiency of the approved method has been studied. When the practical application of the method show that the level of disassembly is directly proportional to the amount of the error square MSE and also has a great effect on the extent of correlation where the recovered image away from the original image in direct proportional with the increased degree of disassembly of the image. It also shows the extent to which it is affected by the image of different types and varieties, where was the highest value of the PSNR (58.0393) in the natural images and the less valuable in x-ray images (56.9295) as shown in table 4.

Image text document can be acquire via scanner or other type of machines which cause sometime a s... more Image text document can be acquire via scanner or other type of machines which cause sometime a skew angle (alignment) on its horizontal direction so many techniques were developed to treat that skew. In This Research a new idea based on Fan filter will be achieved to detect the skew angle of a text image. A fan filter will be designed in different level and size to be suitable to decompose the image to its component with the same number of filter level, then looking for which coefficients hold the highest energy. The skew angle of the text image will be extracted from The fan filter component rank which match the highest energy component in the image slices. The proposed algorithm applied on different type of skew with different images. The difference between the measured angle and the actual angle in the range of ± 1.5o. the proposed algorithm was simulated using Matlab software Version (R2011a).

2019 International Conference on Computing and Information Science and Technology and Their Applications (ICCISTA), 2019

This paper produced a new study of dropping high frequencies from an image decomposed using Conto... more This paper produced a new study of dropping high frequencies from an image decomposed using Contourlet transformation to its coefficients. Studying the effect of decompose level in addition to selecting different high frequency coefficient to be dropped, then reconstruct the image from the remaining coefficients has been investigated. The effect of dropped frequencies on the retrieved image has been evaluated. Performance study has investigated in terms of MSE, PSNR and structural similarity of image. Numerical results have shown less MSE, high PSNR, and high structural similarity for images with large size for the same decomposition levels where structural measure closed to (0.87), (0.99) with less MSE approximately (69), (20) for decomposition levels equal [2]–[4] and [2], [3] respectively. While structural similarity of the image does not highly effected with a change of decomposition levels. The resultant MSE and PSNR show promising results for image high quality in many applications of image processing.

Nucleation and Atmospheric Aerosols, 2023

Object tracking is regarded as one of the significant topics in the scope of computer vision, whi... more Object tracking is regarded as one of the significant topics in the scope of computer vision, which led to rapid development in the practical field through enhancing the reliable tools available in such a field.Recently, the appearance of Artificial Neural Networks (ANNs) resulted in new methods to the identification and recognition of objects. In this research a noble method is suggested for tracking objects inside video files, the method was used to detect cars on the streets through extracting the features of these cars by using You Only Look Ones (YOLO v3) Our method was able to process the extracted features of the cars at detection phase. Also, it exploited the extracted feature in an effective way in order to obtain precise detections. The findings show that our proposed method performs better than the other methods available in the field, as it is able to produce better predictions by using less computation which resulted in reducing the time that such process usually takes. The evaluation results show that our method was able to process an average of 207.6 frames per second to track objects with 67.6% Multi-Object Tracking Accuracy (MOTA) and 89.1% Multi-Object Tracking Precision (MOTP).

2ND INTERNATIONAL CONFERENCE ON MATHEMATICAL TECHNIQUES AND APPLICATIONS: ICMTA2021

Object tracking is regarded as one of the significant topics in the scope of computer vision, whi... more Object tracking is regarded as one of the significant topics in the scope of computer vision, which led to rapid development in the practical field through enhancing the reliable tools available in such a field.Recently, the appearance of Artificial Neural Networks (ANNs) resulted in new methods to the identification and recognition of objects. In this research a noble method is suggested for tracking objects inside video files, the method was used to detect cars on the streets through extracting the features of these cars by using You Only Look Ones (YOLO v3) Our method was able to process the extracted features of the cars at detection phase. Also, it exploited the extracted feature in an effective way in order to obtain precise detections. The findings show that our proposed method performs better than the other methods available in the field, as it is able to produce better predictions by using less computation which resulted in reducing the time that such process usually takes. The evaluation results show that our method was able to process an average of 207.6 frames per second to track objects with 67.6% Multi-Object Tracking Accuracy (MOTA) and 89.1% Multi-Object Tracking Precision (MOTP).

Journal of Al-Qadisiyah for Computer Science and Mathematics

Reliance on computer vision systems in the sports field is one of the very important topics, whic... more Reliance on computer vision systems in the sports field is one of the very important topics, which are of high importance, especially in the arbitration process or evaluating the accuracy of the player’s performance of the movement. It is better to rely on computer vision systems that are more accurate in the arbitration process. In this article, a method was presented to distinguish one of the important movements of the gymnastics player, by relying on deep learning techniques. The dataset was built based on high-quality video clips found on YouTube for tournaments held from the period 2018-2022, due to the absence of The dataset available. This data was divided into three sections: 70% for training, 10% for validation, and 20% for testing. Two models of the convolutional neural network yolov7 and yolov5 were trained, and the results obtained after testing the results of the models show that the seventh version was the best , Recall, Precision and Mean Average Precision criteria we...

The methods of detecting objects and tracking their movements are among the methods that are reli... more The methods of detecting objects and tracking their movements are among the methods that are relied upon in many fields, whether medical or industrial, and others. One of these areas that will rely on deep learning methods in discovering and distinguishing the player's movements is the sports field and is very useful in games in which the player's degree depends on the accuracy of the performance of the movement, such as the gymnastics game, where it was applied to the static ring gymnastics game, where the distinction of movements was discovered the stability in this game is based on a convolutional neural network. Models. The neural network was trained on five of the most important stability movements in this game after creating the data set based on a set of videos of tournaments held in the period from 2016-2022, where an average of 1500 images were obtained for each stability movement, which was divided into 80% for training and 20 % for testing, after training the convolutional neural network model, it was applied to a group of video clips for different tournaments. Many criteria were adopted to measure the efficiency of the model after training and practical application, which showed the efficiency of the proposed system.

Journal of Al-Qadisiyah for Computer Science and Mathematics

Due to progress in the field of deep learning in order to find and track objects through the use ... more Due to progress in the field of deep learning in order to find and track objects through the use of computer vision in the service of large segments of the population, as it was adopted in the field of serving people b with special needs for the sake of dialogue and implementation of many requests in this research, a series of commands for use by people with special needs with speech problems or paralysis, where the ability to use the eye blink is very useful for social communication, were developed. In this research, the orders needed by the target people with special needs were studied, and (11) commands were identified that can be increased according to the intended sample. A table of commands was built depending on the length of the eye blink period. By creating a modified CNN: Convolutional Neural Network structure and training it on 2 different database, deep learning was used to identify and determine if the eye is closed or open (mrlEye2018 and Closed Eye in the Wlid:CEW).It...

Wasit Journal of Pure sciences

The methods of detecting objects and tracking their movements are among the methods that are reli... more The methods of detecting objects and tracking their movements are among the methods that are relied upon in many fields, whether medical or industrial, and others. One of these areas that will rely on deep learning meth-ods in discovering and distinguishing the player's movements is the sports field and is very useful in games in which the player's degree depends on the accura-cy of the performance of the movement, such as the gymnastics game, where it was applied to the static ring gymnastics game, where the distinction of move-ments was discovered the stability in this game is based on a convolutional neu-ral network. Models. The neural network was trained on five of the most im-portant stability movements in this game after creating the data set based on a set of videos of tournaments held in the period from 2016-2022, where an aver-age of 1500 images were obtained for each stability movement, which was di-vided into 80% for training and 20 % for testing, after training t...

Tikrit Journal of Pure Science

Image transformation provide deep meaning about images feature, so many type of image transformat... more Image transformation provide deep meaning about images feature, so many type of image transformation are appear in the last decade years, one of them is curvelet transformation which improve the image processing techniques specially in field of feature extraction. Personal authentication adopt biometric information to be one of the major coefficients in this field. Palm print one of the main approaches for personal identification. So studying the moments extracted from coefficients of curvelet transform of palm print image adopted in order to get high efficient system for personalization systems. Two major phases are constructed in this research to adopt the moments of low frequency coefficient of the curvelet for personal identification. In the first phase a database was built for 50 persons by acquisition nine images for both hands (9 for left hand and 9for right hand). images are acquired and then processed to extract ROI (region of interest) by looking for the palm centroid then...

Journal of University of Anbar for Pure Science

The data hiding is one of the most important subject in field of computer science, so a lot of te... more The data hiding is one of the most important subject in field of computer science, so a lot of technique was developed and modified to satisfy the optimum lend of hiding. In this research the contourlet transformation coefficients were studied to decide which of them are suitable to embed data on it a lot of parameters of the contourlet coefficients can be discussed one of them is the coefficient energy.The research covered most of the suggested events which could be met during the embedding state, one of them the size of the cover in addition to the size of the information were studied. Applying the suggested idea on different type of image with different size (cover image and the message image) shows that the coefficients with low level of energy are suitable to embedded the information, and the retrieved cover and message are so closed to the original one. Key word-image processing; data security; watermarking; contourlet.

JOURNAL OF EDUCATION AND SCIENCE

In this research, the important features of recognizing speakers' identity by extracting features... more In this research, the important features of recognizing speakers' identity by extracting features of speakers voices are used. Voice is considered one of the vital factors adopted algorithm on samples of speakers' voices has been implemented by recording the voice, saving it on a file of the type wav. The voice is treated by using Hamming function to reduce error ratio. Feature of the voice samples have been extracted by taking the spectrum value of the sound signal of all the speakers. Two methods have been used (Ecledian distance and correlation factor) for comparing core of the convolution of the sample of speaker test (a training sample which is being recorded another time) and all the speakers samples who have already been saved on the data base. Applying the algorithm on different voices, ratio of error was very small while matching has increased by the increasing number of speaker. The algorithm has been implemented using mathlab.

IRAOI JOURNAL OF STATISTICAL SCIENCES, Jun 28, 2011

In this research the covariance matrix which was used in so many fields, its eigen values adopted... more In this research the covariance matrix which was used in so many fields, its eigen values adopted to be the main parameters for Latin printed character recognition. The idea was divided in two stages. The first stage is to generate the covariance matrix then, evaluate its eigen values to build main table for the whole latin characters in addition to the numeric values. The second stage is the recognition stage, which achieved to any character that was entered. The applied examples do not register any negative result. So it can be strongly recommended for printed character recognition.

IRAOI JOURNAL OF STATISTICAL SCIENCES, 2013

The 6 th Scientific Conference of the College of Computer Sciences & Mathematics pp [165-180] ] 1... more The 6 th Scientific Conference of the College of Computer Sciences & Mathematics pp [165-180] ] 165 [

In this research it has been compared between two methods used to recognize Arabic character, the... more In this research it has been compared between two methods used to recognize Arabic character, the first based on histogram technique to extract the characteristic for the images of Arabic characters, and used with Kohonen neural network to recognize the Arabic characters. The second method was reliance on the seven invariant moment that extracted from Arabic characters the adopted by back propagation neural network for Arabic character recognition. In the first method, the Kohonen neural network has been trained on vertical histogram values for a set of letters(ث، ح ، د، ر، س، ض، ط، غ، ق، ن), and also it has been tested on a set of noisy letters images, and the results were achieved true, and compatible with the previous one. Also, kohonen network has been trained using horizontal histogram for the same group of characters, and the results had lower closeness from the results that obtained with vertical histogram. The results were achieved true 66%. Again, several type of the backpropagation neural network has been trained in seven invariant moment of the Arabic characters(ا، ب ، ح ، ع ، س ، ظ ، ر ، ذ ، ف، ك) and it has been used several types of the training functions that common used with backpropagation neural network. The results were achieved true 75% . So, it is discern previously that depending on vertical histogram values with kohonen neural network gives higher closeness from horizontal histogram vales with kohonen neural network and seven invariant moment with back propagation neural network. In addition, when depending on the two histograms together, it was gives bad results with other neural network as well as kohonen neural network.

AL-Rafidain Journal of Computer Sciences and Mathematics, 2021

With the rapid growth of computer usage to extract the required knowledge from a huge amount of i... more With the rapid growth of computer usage to extract the required knowledge from a huge amount of information, such as a video file, significant attention has been brought towards multi-object detection and tracking. Artificial Neural Networks (ANNs) have shown outstanding performance in multi-object detection, especially the Faster R-CNN network. In this study, a new method is proposed for multi-object tracking based on descriptors generated by a neural network that is embedded in the Faster R-CNN. This embedding allows the proposed method to directly output a descriptor for each object detected by the Faster R-CNN, based on the features detected by the Faster R-CNN to detect the object. The use of these features allows the proposed method to output accurate values rapidly, as these features are already computed for the detection and have been able to provide outstanding performance in the detection stage. The descriptors that are collected from the proposed method are then clustered into a number of clusters equal to the number of objects detected in the first frame of the video. Then, for further frames, the number of clusters is increased until the distance between the centroid of the newly created cluster and the nearest centroid is less than the average distance among the centroids. Newly added clusters are considered for new objects, whereas older ones are kept in case the object reappears in the video. The proposed method is evaluated using the UA-DETRAC (University at Albany Detection and Tracking) dataset and has been able to achieve 64.8% MOTA and 83.6% MOTP, with a processing speed of 127.3 frames per second.

AL-Rafidain Journal of Computer Sciences and Mathematics, 2021

The interest in coding was very high because it is widely relied on in the security of correspond... more The interest in coding was very high because it is widely relied on in the security of correspondence and in the security of information in addition to the need to rely on it in the storage of data because it leads to a pressure in the volume of information when storing it. In this research, image transformation was used to encode gray or color images by adopting parameters elected from contourlet transformations for image. The color images are acquired into the algorithm, to be converted into three slices (the main colors of the image), to be disassembled into their coefficients through contourlet transformations and then some high frequencies in addition to the low frequency are elected in order to reconstruct the image again. The election of low frequencies with a small portion of the high frequencies has led to bury some unnecessary information from the image components. The performance efficiency of the proposed method was measured by MSE and PSNR criteria to see the extent of the discrepancy between the original image and the recovered image when adopting different degrees of disassembly level, in addition, the extent to which the image type affects the performance efficiency of the approved method has been studied. When the practical application of the method show that the level of disassembly is directly proportional to the amount of the error square MSE and also has a great effect on the extent of correlation where the recovered image away from the original image in direct proportional with the increased degree of disassembly of the image. It also shows the extent to which it is affected by the image of different types and varieties, where was the highest value of the PSNR (58.0393) in the natural images and the less valuable in x-ray images (56.9295) as shown in table 4.

Image text document can be acquire via scanner or other type of machines which cause sometime a s... more Image text document can be acquire via scanner or other type of machines which cause sometime a skew angle (alignment) on its horizontal direction so many techniques were developed to treat that skew. In This Research a new idea based on Fan filter will be achieved to detect the skew angle of a text image. A fan filter will be designed in different level and size to be suitable to decompose the image to its component with the same number of filter level, then looking for which coefficients hold the highest energy. The skew angle of the text image will be extracted from The fan filter component rank which match the highest energy component in the image slices. The proposed algorithm applied on different type of skew with different images. The difference between the measured angle and the actual angle in the range of ± 1.5o. the proposed algorithm was simulated using Matlab software Version (R2011a).

2019 International Conference on Computing and Information Science and Technology and Their Applications (ICCISTA), 2019

This paper produced a new study of dropping high frequencies from an image decomposed using Conto... more This paper produced a new study of dropping high frequencies from an image decomposed using Contourlet transformation to its coefficients. Studying the effect of decompose level in addition to selecting different high frequency coefficient to be dropped, then reconstruct the image from the remaining coefficients has been investigated. The effect of dropped frequencies on the retrieved image has been evaluated. Performance study has investigated in terms of MSE, PSNR and structural similarity of image. Numerical results have shown less MSE, high PSNR, and high structural similarity for images with large size for the same decomposition levels where structural measure closed to (0.87), (0.99) with less MSE approximately (69), (20) for decomposition levels equal [2]–[4] and [2], [3] respectively. While structural similarity of the image does not highly effected with a change of decomposition levels. The resultant MSE and PSNR show promising results for image high quality in many applications of image processing.