Topic
Histogram of oriented gradients
About: Histogram of oriented gradients is a research topic. Over the lifetime, 2037 publications have been published within this topic receiving 55881 citations. The topic is also known as: HOG.
Papers published on a yearly basis
Papers
More filters
••
01 Jun 2017
TL;DR: The most aggressive simplification is the removal of the bin interpolation step in the algorithm, which does not affect the classification performance, but significantly reduces the FPGA area, which finally allows more parallelism and hence a faster processing speed of a full image.
Abstract: In this paper we propose and investigate some simplifications to the original Histogram of Oriented Gradients (HOG) algorithm, in order to allow a more efficient hardware implementation, while keeping the overall classification accuracy. The most aggressive simplification is the removal of the bin interpolation step in the algorithm, which does not affect the classification performance, but significantly reduces the FPGA area, which finally allows more parallelism and hence a faster processing speed of a full image.
9 citations
••
01 Jan 2019TL;DR: This paper describes object classification using artificial neural network with back-propagation as a Feed-Forward network and this neural network has been used to classify different categories of objects based on the features extracted and trained.
Abstract: The proposed work summarizes approach of using histogram of gradients as descriptor which are taken as training features for the neural network This paper describes object classification using artificial neural network with back-propagation as a Feed-Forward network HOG features were extracted from the images to pass on to this feed-forward network and this neural network has been used to classify different categories of objects based on the features extracted and trained The converging condition is determined and analyzed for the designed approach The experimental neural network comprises of 64 neurons in the input layer and 16 neurons in the hidden layer and the output layer has 4 neurons, which is the number of classes The accuracy for training as well as testing will be discussed and provided in a tabular form up to 3500 epochs All experimental results are shown in form of graphs and tables
9 citations
••
TL;DR: A novel feature descriptor based on the integration of magnitude and orientation information of optical flow and histogram of oriented gradients which gives an efficient and robust feature vector for the recognition of human activities for real-world environment is introduced.
Abstract: Human activity recognition from video sequences has emerged recently as pivotal research area due to its importance in a large number of applications such as real-time surveillance monitoring, heal...
9 citations
••
25 Jul 2013TL;DR: The kernel function for SVM classifier, suitable for individual object classifier is analysed based upon ROC-AUC parameter and is validated with performance metrics like precision, recall and accuracy.
Abstract: Multiclass object detection is considered for detecting different object classes in a cluttered environment. Traditional approaches require applying a battery of different classifiers to the image with a large number of complex features used to detect the objects. Specialized detectors usually excel in performance, while the class-specific features increase detection accuracy, but at the expense of complexity. In this paper, an efficient method of human face and car detection using cascaded structure of independent object classifiers is proposed. The approach is based on background elimination using statistical features, followed by foreground detection using Principal component analysis (PCA) and Histogram of Gradients (HoG) with SVM classifier. For detecting the object of interest from the image, the system primarily filters the potential object area by analyzing the local histogram distribution. After background elimination, the trained classifier detects foreground using higher order parameters like PCA for human faces and HOG for cars. In this paper, the kernel function for SVM classifier, suitable for individual object classifier is analysed based upon ROC-AUC parameter. The proposed system is implemented in Matlab. The system is validated with performance metrics like precision, recall and accuracy.
9 citations
••
18 Mar 2016TL;DR: This paper presents a copy move forgery detection algorithm which detects and localizes the forgery with reasonably good accuracy and in less time compared to other state-of-the-art methods.
Abstract: Image forgery detection is always a challenging task especially when it comes to the blind image forgery detection techniques which do not contain any water marks or signatures. Copy move forgery is a popular blind image forgery detection technique which is widely used by forgers for performing easy and efficient forgery. A copy-move forgery followed by rotation and scaling of the forged part is even more challenging to detect. In this paper, we present a copy move forgery detection algorithm which detects and localizes the forgery with reasonably good accuracy and in less time compared to other state-of-the-art methods. This method uses Harris Corner points to detect regions of interest and identifies a feature vector around this corner points using the Histogram of Oriented Gradients (HOG) descriptor. A comparison of the feature vectors is performed in the entire feature space to find a suitable match by using Sum of Squared Differences (SSD) and Nearest Neighbour Distance Ratio (NDR). The outlier matches in the previous step is eliminated by using RANSAC. The results are evaluated based on various performance measures like True Positive Rate (TPR), False Positive Rate (FPR), Precision and Recall. Datasets proposed by Adrizzone et al. and CoMoFoD(small) are used for evaluation. The results achieved show the efficiency of our method against moderate rotation and scaling with a reasonably good True Positive Rate.
9 citations