Pengenalan Wajah Menggunakan Pembelajaran Mesin Berdasarkan Ekstraksi Fitur Pada Gambar Wajah Berkualitas Rendah

Wajah merupakan bagian dari sistem biometric dimana wajah manusia memiliki bentuk dan karakteristik yang berbeda antara satu dengan lainnya sehingga wajah dapat dijadikan sebagai alternatif pengamanan suatu sistem. Proses pengenalan wajah didasarkan pada proses pencocokan dan perbandingan citra yang dimasukan dengan citra yang telah tersimpan di database. Akan tetapi pengenalan wajah menjadi permasalahan yang cukup menantang dikarenakan illuminasi, pose dan ekspresi wajah serta kualitas citra. Oleh sebab itu pada penelitian ini bertujuan untuk melakukan pengenalan wajah dengan menggunakan metode machine learning seperti Logistic Regression (LR), Linear Discriminant Analysis (LDA), Decision Tree Classifier, Random Forest Classifier (RF), Gaussian NB, K Neighbors Classifier (KNN) dan Support Vector Machine (SVM) dan beberapa metode ekstraksi fitur Hu-Moment, HOG dan Haralick pada dataset Yale Face. Berdasarkan pengujian yang dilakukan metode ekstraksi fitur gabungan Hu-Moment, HOG dan Haralick dengan algoritma Linear Discriminant Analysis (LDA) menghasilkan nilai akurasi tertinggi sebesar 79,71% dibandingkan dengan metode ekstraksi fitur dan algoritma klasifikasi lainnya.

Bayesian Pixel Density Estimation Modeling to Detect Human Sperm Sample Image Based on Sperm Head Shape

Currently, there is a problem of the difficulty in classifying human sperm head sample images using different databases and measuring the accuracy of several different datasets. This study proposes a Bayesian Density Estimation-based model for detecting human sperm heads with four classification labels, namely, normal, tapered, pyriform, and small or amorphous. This model was applied to three kinds of datasets to detect the level of pixel density in images containing normal human sperm head samples. Experimental results and computational accuracy are also presented. As a method, this study labeled each human sperm head based on three shape descriptors using the formulas of Hu moment, Zernike moment, and Fourier descriptor. Each descriptor was also tested in the experiment. There was an increased accuracy that reached 90% after the model was applied to the three datasets. The Bayesian Density Estimation model could classify images containing human sperm head samples. The correct classification level was obtained when the human sperm head was detected by combining Bayesian + Hu moment with an accuracy rate of up to 90% which could detect normal human sperm heads. It is concluded that the proposed model can detect and classify images containing human sperm head objects. This model can increase accuracy, so it is very appropriate to be applied in the medical field

Rate of craniometric change following suture release in patients with metopic and sagittal craniosynostosis

OBJECTIVE Craniosynostosis is a congenital disorder resulting from the premature fusion of cranial sutures in the infant skull. This condition results in significant cosmetic deformity and can impede neurodevelopment, if left untreated. Currently, rates of craniometric change following minimally invasive surgery have only been examined for sagittal craniosynostosis. A better understanding of postoperative skull adaptations in other craniosynostosis subtypes is needed to objectively categorize surgical outcomes and guide length of cranial orthosis therapy. METHODS Eleven patients with sagittal and 8 with metopic craniosynostosis treated using endoscopic strip craniectomy and postoperative helmet orthoses were retrospectively reviewed. Using semiautomated image analysis of top-down orthogonal 2D photographs, the following craniometrics were recorded before surgery and at postoperative visits: cephalic index (CI), cranial vault asymmetry index (CVAI), anterior arc angle (AAA), posterior arc angle (PAA), anterior-middle width ratio (AMWR), anterior-posterior width ratio (APWR), left-right height ratio (LRHR), sagittal Hu moment (Sag-Hu), and brachycephaly Hu moment (Brachy-Hu). These craniometrics were then normalized to photograph-based measurements of normocephalic patients and the rates of change between metopic and sagittal craniosynostoses were compared. RESULTS Patients with sagittal craniosynostosis exhibited significantly lower CI, lower PAA, higher AMWR, higher APWR, lower Sag-Hu, and higher Brachy-Hu preoperatively compared to patients with normocephalic craniosynostosis. Patients with metopic craniosynostosis exhibited lower AAA and AMWR preoperatively compared to normocephalic subjects. Sagittal and metopic patients had a rapid initial change in normalized CI or AAA, respectively. Craniometric rates of change that significantly differed between metopic and sagittal patients were found in AAA (p < 0.001), AMWR (p < 0.001), and APWR (p < 0.0001). Metopic patients had a prolonged AAA change with a significantly different rate of change up to 6 months postoperatively (median at 3 months = 0.027 normalized units/day, median at 6 months = 0.017 normalized units/day, and median at > 6 months = 0.007 normalized units/day), while sagittal CI rate of change at these time points was not significantly different. CONCLUSIONS Patients with metopic craniosynostosis have a prolonged rate of change compared to patients with sagittal craniosynostosis and may benefit from longer helmet use and extended postoperative follow-up. Categorizing craniometric changes for other craniosynostosis subtypes will be important for evaluating current treatment guidelines.

FGFF Descriptor and Modified Hu Moment-Based Hand Gesture Recognition

Gesture recognition has been studied for decades and still remains an open problem. One important reason is that the features representing those gestures are not sufficient, which may lead to poor performance and weak robustness. Therefore, this work aims at a comprehensive and discriminative feature for hand gesture recognition. Here, a distinctive Fingertip Gradient orientation with Finger Fourier (FGFF) descriptor and modified Hu moments are suggested on the platform of a Kinect sensor. Firstly, two algorithms are designed to extract the fingertip-emphasized features, including palm center, fingertips, and their gradient orientations, followed by the finger-emphasized Fourier descriptor to construct the FGFF descriptors. Then, the modified Hu moment invariants with much lower exponents are discussed to encode contour-emphasized structure in the hand region. Finally, a weighted AdaBoost classifier is built based on finger-earth mover’s distance and SVM models to realize the hand gesture recognition. Extensive experiments on a ten-gesture dataset were carried out and compared the proposed algorithm with three benchmark methods to validate its performance. Encouraging results were obtained considering recognition accuracy and efficiency.

Method of extracting design knowledge from 3D models

The mathematical description of extracting design knowledge from 3D-models and methods for comparing the obtained knowledge, used in the search for geometrically similar mechanical engineering products, are considered. Keywords: 3D-model, design knowledge, deformation method, Hu-moment, convolutional neural [email protected]

Video interframe forgery detection: Classification, technique & new dataset

Multimedia communication as well as other related innovations are gaining tremendous growth in the modern technological era. Even though digital content has traditionally proved to be a piece of legitimate evidence. But the latest technologies have lessened this trust, as a variety of video editing tools have been developed to modify the original video. Therefore, in order to resolve this problem, a new technique has been proposed for the detection of duplicate video sequences. The present paper utilizes gray values to extract Hu moment features in the current frame. These features are further used for classification of video as authentic or forged. Afterwards there was also need to validate the proposed technique using training and test dataset. But the scarcity of training and test datasets, however, is indeed one of the key problems to validate the effectiveness of video tampering detection techniques. In this perspective, the Video Forensics Library for Frame Duplication (VLFD) dataset has been introduced for frame duplication detection purposes. The proposed dataset is made of 210 native videos, in Ultra-HD and Full-HD resolution, captured with different cameras. Every video is 6 to 15 seconds in length and runs at 30 frames per second. All the recordings have been acquired in three different scenarios (indoor, outdoor, nature) and in landscape mode(s). VLFD includes both authentic and manipulated video files. This dataset has been created as an initial repository for manipulated video and enhanced with new features and new techniques in future.

Klasifikasi Buah Segar dan Busuk Menggunakan Ekstraksi Fitur Hu-Moment , Haralick dan Histogram

Seiring dengan perkembangan zaman, teknologi berkembang dengan pesat saat ini. Dengan perkembangan teknologi sekarang ini memudahkan semua orang mengakses apa saja. Banyak teknologi yang sudah ditemukan salah satunya adalah pengolahan citra digital, Identifikasi pada sebuah citra sudah lama dikembangkan salah satunya dengan membedakan tekstur pada citra tersebut. Tekstur citra dapat dibedakan oleh kerapatan, keseragaman, kekasaran dan keteraturan dari citra yang diteliti. Agriculture saat ini sedang ramai di bahas khususnya di indonesia, banyak sekali penelitian yang di lakukan dalam sektor pertanian guna memajukan sektor pertanian itu sendiri. Dalam penelitian kali ini yaitu ekstraksi fitur menggunakan Hu-moment, Haralick dan Histogram dan klasifikasi menggunakan algoritma Random Forest. Peneliti mencoba mengklasifikasi buah-buahan segar atau busuk, dengan algoritma yang digunakan yaitu algoritma Random Forest, penelitian ini mendapatkan akurasi yang sangat tinggi yakni 99.6% sangat baik sekali. Namun guna memperbaharui penelitian bisa mencoba beberapa fitur dan algorithma yang berbeda agar mendapatkan perbandingan atau hasil yang lebih maksimal.Kata kunci: ekstraksi fitur, Hu-moment Haralick dan Histogram, Random Forest.