Boosted of Haar-like Features and Local Binary Pattern Based Face Detection (original) (raw)
Related papers
Development of real time Face detection system using Haar like features and Adaboost algorithm
2010
Human face detection is an active area of research covering several disciplines such as image processing, pattern recognition and computer vision. This paper describes a face detection framework that is capable of processing input images pretty swiftly while achieving high detection rates. The existing methods for face detection can be divided into image based methods and feature based methods. The developed system is intermediary of these two, using a hybrid method comprising boosting algorithm and a hyper plane to train a classifier which is capable of processing images rapidly while having high detection rates. Using the response of simple Haar-based features used by Viola and Jones [1], AdaBoost algorithm and an additional hyper plane classifier, the presented face detection system is developed. This system is further modified by some intuitive noble heuristics. A set of experiments in the domain of face detection is presented. The system yields face detection performance comparable to the best previous systems
Face detection based on multi-block lbp representation
Advances in Biometrics, 2007
Effective and real-time face detection has been made possible by using the method of rectangle Haar-like features with AdaBoost learning since Viola and Jones' work . In this paper, we present the use of a new set of distinctive rectangle features, called Multi-block Local Binary Patterns (MB-LBP), for face detection. The MB-LBP encodes rectangular regions' intensities by local binary pattern operator, and the resulting binary patterns can describe diverse local structures of images. Based on the MB-LBP features, a boosting-based learning method is developed to achieve the goal of face detection. To deal with the non-metric feature value of MB-LBP features, the boosting algorithm uses multibranch regression tree as its weak classifiers. The experiments show the weak classifiers based on MB-LBP are more discriminative than Haar-like features and original LBP features. Given the same number of features, the proposed face detector illustrates 15% higher correct rate at a given false alarm rate of 0.001 than haar-like feature and 8% higher than original LBP feature. This indicates that MB-LBP features can capture more information about the image structure and show more distinctive performance than traditional haar-like features, which simply measure the differences between rectangles. Another advantage of MB-LBP feature is its smaller feature set, this makes much less training time.
Fast training and selection of Haar features using statistics in boosting-based face detection
International Conference on Computer Vision, 2007
Training a cascade-based face detector using boosting and Haar features is computationally expensive, often requiring weeks on single CPU machines. The bottleneck is at training and selecting Haar features for a single weak classifier, currently in minutes. Traditional techniques for training a weak classifier usually run in O(N T log N ), with N examples (approximately 10,000), and T features (approximately 40,000). We present a method to train a weak classifier in time O(N d 2 + T ), where d is the number of pixels of the probed image sub-window (usually from 350 to 500), by using only the statistics of the weighted input data. Experimental results revealed a significantly reduced training time of a weak classifier to the order of seconds. In particular, this method suffers very minimal immerse in training time with very large increases in members of Haar features, enjoying a significant gain in accuracy, even with reduced training time.
TRAINING ANALYSIS OF HAAR-CLASSIFIERS FOR FACE DETECTION
IAEME PUBLICATION, 2021
Haar features have been used in most of the works in literature as key components in the task of object as well as face detection. Training process of Haar features is an important step in the development of overall face detection system. A number of features can be observed in the training process of Haar features. In this work, we have done investigation in the training parameters during AdaBoost based machine learning of Haar features. Based on the studying of parameters during training process, efficient learners can be selected from a large pool of available features which are further cascaded to make the strong classifiers. Modification in face detection process by application of scaling of detector window by changing the base size of training samples and efficient detection technique has been proposed.
A Face Detection Using Haar Like Feature Algorithm
This paper utilizes another face discovery strategy dependent on Haar-Like element. New Haar-Like component is an expansion of the Haar-Like include premise. This article utilize four new Haar-Like include, and these highlights with existing Haar-Like element are input Adaboost classifier together to select component, at long last developed characterization execution and incredible course classifier for face location. After location tests we can it couldn't be any more obvious, the calculation can show signs of improvement results analyzed with other conventional face recognition classifiers like Haar-Like.
Face Detection and Recognition Using Haar Classifier and LBP Histogram
International Journal of Advanced Research in Computer Science, 2018
Facial recognition technology is the process for identifying or verifying a face from digital images. The need for face recognition has been of real importance with the development of modern society. Detection and recognition of faces has been on the rise worldwide owing the requirement for security for economic transactions, authorization, national safety and security and other important factors. The technology comprises of face detection, database creation and face recognition. This paper presents a new approach of face identification using LBP method and Haar-like features. The first step is face detection which is done using Haar cascade classifier. After detection, a face is saved in the database. Then the faces from the database are passed through the face recognition algorithm. The Local Binary Pattern Histogram (LBPH) method is used for face recognition. The performance of face detection can be seen to produce maximum error of 1.6%, 2.1% and 0.8% in case of Real-Time video, image file and video file respectively which may be considered accurate. The recognition algorithm produces maximum error of 0.4% which may be considered accurate as well.
Face Detection System On AdaBoost Algorithm Using Haar Classifiers
2012
This paper presents an architecture for face detection based system on AdaBoost algorithm using Haar features. We describe here design techniques including image scaling, integral image generation, pipelined processing as well as classifier, and parallel processing multiple classifiers to accelerate the processing speed of the face detection system. Also we discuss the optimization of the proposed architecture which can be scalable for configurable devices with variable resources. The proposed architecture for face detection has been designed using Verilog HDL and implemented in Modelsim. Its performance has been measured and compared with an equivalent hardware implementation. We show about 35 time's increase of system performance over the equivalent hardware implementation.
REAL TIME FACE DETECTION USING LOCAL BINARY PATTERN FEATURES
Face Detection and Recognition is an important area in the field of substantiation. Maintenance of records of students along with monitoring of class attendance is an area of administration that requires significant amount of time and efforts for management. Automated Attendance Management System performs the daily activities of attendance analysis, for which face recognition is an important aspect. The prevalent technique and methodologies for detecting and recognizing face like PCA-LDA, etc fail to overcome issues such as scaling, pose, illumination, variations, rotation, and occlusions. The proposed system provides features such as detection of faces, extraction of the features, detection of extracted features, analysis of students' attendance and monthly attendance report generation. The proposed system integrates techniques such as image contrasts, integral images, Ada-Boost, Haar-like features and cascading classifier for feature detection. Faces are recognized using advanced LBP using the database that contains images of students and is used to recognize student using the captured image. Better accuracy is attained in results and the system takes into account the changes that occurs in the face over the period of time.
6th Pacific-Rim Symposium on Image and Video Technology; PSIVT, 2014
This paper addresses the problem of detecting human faces in noisy images. We propose a method that includes a denoising preprocessing step, and a new face detection approach based on a novel extension of Haar-like features. Preprocessing of the input images is focused on the removal of different types of noise while preserving the phase data. For the face detection process, we introduce the concept of global and dynamic global Haar-like features, which are complementary to the well known classical Haar-like features. Matching dynamic global Haar-like features is faster than that of the traditional approach. Also, it does not increase the computational burden in the learning process. Experimental results obtained using images from the MIT-CMU dataset are promising in terms of detection rate and the false alarm rate in comparison with other competing algorithms. 1. Speeding up the learning process. 2. Speeding up the face detection. 3. Defining a better trade-off between detection rate and false-positive rate. 4. Combining the three mentioned criteria. For example, heuristic methods trying to improve the detection speed [13], or different version of AdaBoost like. Float boost [5], ChainBoost [20], costsensitive boosting [9, 1], KLBoost [7], FCBoost [16], or RCECBoost [15] that aim at speeding up the AdaBoost convergence, or at improving the final performance of the detector.
ADAPTIVE WEIGHT ANALYSIS FOR LEARNING OF A HAAR-CLASSIFIERS BASED FACE DETECTION CASCADE
IAEME PUBLICATION, 2021
Most of the works in the field of object detection since past two decades have chosen Haar-features for training and building of the detection cascade because of their simplicity and versatility. The Haar-features based classifiers have been used not only for the face detection but also for the detection of facial landmarks and features extraction. A number of researchers have also used Haar-features for the task of face quality estimation as well as for quality enhancement of low quality faces. Therefore because of the holistic nature of Haar-features and their simplicity, Haar-features were selected for the work in the field of face detection. As per AdaBoost, the weights of all of the example images are updated during an iteration of processing a feature, which is adaptive in nature. In this work we have done the analysis of variations of adaptive weights of various features. As per analysis it has been found that out of total 6976 training images, 93.88% images have undergone decrease in their adaptive weight while remaining 427 images only have undergone raise in weight during the training process, which shows the efficiency of the AdaBoost algorithm in training.