Non-maximum Suppression Algorithm Research Articles

Pig Detection Algorithm Based on Sliding Windows and PCA Convolution

An accurate and rapid pig detection algorithm based on video image processing technology can be helpful to identify abnormal pigs and take timely measures to reduce the incidence of diseases. In order to solve the problems of low computational efficiency and low precision in pig detection algorithm based on sliding windows, this paper proposed a simple and efficient pig detection algorithm. A two-level support vector machine model was trained to calculate the probabilities of sliding windows by using gradient and gray distribution features of pigs. The principal component analysis convolution kernels were trained to extract foreground and background features of pig images. The support vector machine was used to classify sliding windows to obtain windows where pigs were located, and the non-maximum suppression algorithm was used to eliminate redundant windows to complete the target detection. The experiments showed that the proposed algorithm blending gradient and gray distribution features had a higher recall rate than the BING algorithm. The recall rate was up to 99.21% using 500 windows. The classification accuracy of sliding windows in this paper was 95.21%, which was higher than that of the PCANet. By calculating the omission detection rate, the misdetection rate, and the average detection time, it can be seen that in the detection methods of the proposed algorithm, BING + PCANet, faster rcnn and yolo, the performance of the proposed algorithm was optimal.

An Improved Faster R-CNN for Small Object Detection

With the increase of training data and the improvement of machine performance, the object detection method based on convolutional neural network (CNN) has become the mainstream algorithm in field of the current object detection. However, due to the complex background, occlusion and low resolution, there are still problems of small object detection. In this paper, we propose an improved algorithm based on faster region-based CNN (Faster R-CNN) for small object detection. Using the two-stage detection idea, in the positioning stage, we propose an improved loss function based on intersection over Union (IoU) for bounding box regression, and use bilinear interpolation to improve the regions of interest (RoI) pooling operation to solve the problem of positioning deviation, in the recognition stage, we use the multi-scale convolution feature fusion to make the feature map contain more information, and use the improved non-maximum suppression (NMS) algorithm to avoid loss of overlapping objects. The results show that the proposed algorithm has good performance on traffic signs whose resolution is in the range of (0, 32], the algorithm's recall rate reaches 90%, and the accuracy rate reaches 87%. Detection performance is significantly better than Faster R- CNN. Therefore, our algorithm is an effective way to detect small objects.

IdentifyNet for Non-Maximum Suppression

Two-stage object detectors have achieved great success in recent years. However, recent work mostly focuses on optimizing loss functions or learning multi-level feature representation, while introducing additional homogeneous task to improve detection has been under-explored. In this paper, a novel framework named as IdentifyNet is proposed, which incorporates an additional identification task to enhance the feature learning of region proposals. Specifically, besides classification and bounding box regression, the proposed IdentifyNet further learns to predict whether two different region proposals belong to the same object, thus forcing the network to learn more informative and representative features for different proposals, especially for those from the same object class. Moreover, current detectors apply greedy non-maximum suppression to remove duplicated boxes whenever their Intersaction-over-Union (IoU) exceeds a preset threshold, which would fail when two boxes largely overlap with each other while belonging to two different objects of the same class. To overcome this, we further propose a novel decode non-maximum suppression algorithm by taking advantage of the predicted identity information of different proposals from the identification task. Extensive experiments on PASCAL VOC 2007, VOC 2012 well demonstrate the proposed method can greatly improve detection performance.

Multiscale Block Fusion Object Detection Method for Large-Scale High-Resolution Remote Sensing Imagery

Object detection in high-spatial-resolution remote sensing images (HSRIs) is an import part of the automatic extraction and understanding of image information in high-resolution earth observation systems. Regarding how to achieve optimal block object detection for large-scale HSRIs, this paper proposes a multiscale block fusion object detection method for large-scale HSRIs. First, the objects in large-scale HSRIs are detected using different block scales, and the average precision (AP) of the different object detection results is counted at different block scales. Then, according to the statistical information, the image block scales corresponding to the optimal AP value of the different objects are obtained. Finally, a soft non-maximum suppression algorithm is used to fuse the image block scale detection results corresponding to the optimal AP values of the different objects, to obtain the object detection results of the large-scale HSRIs. The experimental results confirm that the proposed method outperforms all other single-scale image block detection methods and provides acceptable object detection results in large-scale HSRIs.

Detection of palm oil bud rot employing artificial vision

This paper describes the development of a system capable of capturing, processing, and analysing images of palm oil plantations in order to detect and identify bud rot disease. The process begins with the capturing of images using the DJI Phantom 4 UAV, which is first configured for a flight plan of the desired area and altitude. The resulting images are processed by photogrammetry software to create orthomosaics. The developed algorithm uses the grayscale of the generated orthomosaic and identifies the palms affected by bud rot. This is accomplished using the sliding-window method, by which smaller samples of the image are used and evaluated independently. Each sample is then extracted with a ULBP feature vector that numerically represents the texture of the image and is classified by means of a previously trained logistic regression model, allowing the recognition of positive cases of the disease. Possible positive cases are further distinguished using the non-maximum suppression algorithm. The system was tested with different images than the images used for training and for establishing the set point. As a result, the system showed a 92% precision and a 96% of sensitivity for bud rot disease detection. These results are satisfactory in terms of detecting bud rot disease using a low-cost system.

Feature proposal model on multidimensional data clustering and its application

Unsupervised learning based on clustering is an important field of artificial intelligence. In recent years, most studies focused on improving the performance of clustering algorithms by balancing the number of samples in each class or removing the noise features by iterative optimization. However, these methods are not suitable for some practical classification tasks, such as image segmentation and salient region proposals. Due to pixels or superpixels belonging to different objects in different images are not balanced, and it is difficult to use a limited number of features to describe classifying different objects, these algorithms cannot achieve high-performance classification in these tasks. In this paper, we put forward a feature proposal (FeatPro) model that can dynamically select useful features for different tasks. Furthermore, we design a feature non-maximum suppression (FNMS) algorithm, in order to achieve high performance classification by using as few features as possible. Experiments show that our method obtains competitive results on standard classification datasets and in the application of salient object segmentation.

An improved method for SIFT-based copy–move forgery detection using non-maximum value suppression and optimized J-Linkage

In looking to improve the detection performance of the keypoint-based method involving smooth tampered regions, there are three problems to be addressed, namely the nonuniform distribution of the keypoints, the discriminative power of low contrast keypoints, and the high computational cost of clustering. In this study, the classical implementation framework of the keypoint-based method is improved by introducing new techniques and algorithms in order to overcome these problems. First, to acquire uniformly distributed keypoints in the test image, we propose a new solution of selecting the keypoints by region instead of contrast. To this end, we first separate the keypoint detection and selection processes. After obtaining all discernible keypoints, we adapt the non-maximum value suppression algorithm to select keypoints by combining the contrast and density of each keypoint. Second, we apply the opponent scale-invariant feature transform descriptor to enhance the discriminative power of keypoints by adding color information. Finally, to alleviate the computational cost of clustering, we optimize the J-Linkage algorithm by altering the method of computing initial clusters and affine transformation hypotheses. For this purpose, we propose the matched pair grouping algorithm that can obtain a smaller number of initial clusters by utilizing the correspondence between the superpixels in the original and duplicated regions. Experiments performed on three representative datasets confirm that the proposed method can significantly improve the detection performance in smooth tampered regions, and considerably reduce the clustering time in the case of a mass of matched pairs, compared with the state-of-the-art methods.

Accurate Object Localization in Remote Sensing Images Based on Convolutional Neural Networks

In this paper, we focus on tackling the problem of automatic accurate localization of detected objects in high-resolution remote sensing images. The two major problems for object localization in remote sensing images caused by the complex context information such images contain are achieving generalizability of the features used to describe objects and achieving accurate object locations. To address these challenges, we propose a new object localization framework, which can be divided into three processes: region proposal, classification, and accurate object localization process. First, a region proposal method is used to generate candidate regions with the aim of detecting all objects of interest within these images. Then, generic image features from a local image corresponding to each region proposal are extracted by a combination model of 2-D reduction convolutional neural networks (CNNs). Finally, to improve the location accuracy, we propose an unsupervised score-based bounding box regression (USB-BBR) algorithm, combined with a nonmaximum suppression algorithm to optimize the bounding boxes of regions that detected as objects. Experiments show that the dimension-reduction model performs better than the retrained and fine-tuned models and the detection precision of the combined CNN model is much higher than that of any single model. Also our proposed USB-BBR algorithm can more accurately locate objects within an image. Compared with traditional features extraction methods, such as elliptic Fourier transform-based histogram of oriented gradients and local binary pattern histogram Fourier, our proposed localization framework shows robustness when dealing with different complex backgrounds.

An Effective Method Based on ACF for Aircraft Detection in Remote Sensing Images

Detecting artificial targets, such as aircraft, in satellite images is significant in military and civil applications. Although the performance has improved with the use of more complicated features and better learning methods, effectively handling aircraft with variations of type, pose, and size is still very challenging. To solve this problem, we propose a multiscale sliding-window framework based on aggregate channel features, well-designed features that contain rich information. We also employ a fast feature pyramids algorithm to accelerate multiscale aircraft detection. In this framework, features are trained by Cascade AdaBoost including multiple rounds of bootstrapping that leads to improved overall accuracy. A two-step nonmaximum suppression algorithm is carefully designed based on a given set of detections. Our method shows a competitive performance on the QuickBird images of 0.6 m resolution.

Development of an aerial counting system in oil palm plantations

This paper proposes the development of a counting aerial system capable of capturing, process and analyzing images of an oil palm plantation to register the number of cultivated palms. It begins with a study of the available UAV technologies to define the most appropriate model according to the project needs. As result, a DJI Phantom 2 Vision+ is used to capture pictures that are processed by a photogrammetry software to create orthomosaics from the areas of interest, which are handled by the developed software to calculate the number of palms contained in them. The implemented algorithm uses a sliding window technique in image pyramids to generate candidate windows, an LBP descriptor to model the texture of the picture, a logistic regression model to classify the windows and a non-maximum suppression algorithm to refine the decision. The system was tested in different images than the ones used for training and for establishing the set point. As result, the system showed a 95.34% detection rate with a 97.83% precision in mature palms and a 79.26% detection rate with a 97.53% precision in young palms giving an FI score of 0.97 for mature palms and 0.87 for the small ones. The results are satisfactory getting the census and high-quality images from which is possible to get more information from the area of interest. All this, achieved through a low-cost system capable of work even in cloudy conditions.

EXTENSION AND EVALUATION OF THE AGAST FEATURE DETECTOR

Vision-aided inertial navigation is a navigation method which combines inertial navigation with computer vision techniques. It can provide a six degrees of freedom navigation solution from passive measurements without external referencing (e.g. GPS). Thus, it can operate in unknown environments without any prior knowledge. Such a system, called IPS (Integrated Positioning System) is developed by the German Aerospace Center (DLR). <br><br> For optical navigation applications, a reliable and efficient feature detector is a crucial component. With the publication of AGAST, a new feature detector has been presented, which is faster than other feature detectors. To apply AGAST to optical navigation applications, we propose several methods to improve its performance. Based on a new non-maximum suppression algorithm, automatic threshold adaption algorithm in combination with an image split method, the optimized AGAST provides higher reliability and efficiency than the original implementation using the Kanade Lucas Tomasi (KLT) feature detector. Finally, we compare the performance of the optimized AGAST with the KLT feature detector in the context of IPS. The presented approach is tested using real data from typical indoor scenes, evaluated on the accuracy of the navigation solution. The comparison demonstrates a significant performance improvement achieved by the optimized AGAST.

EXTENSION AND EVALUATION OF THE AGAST FEATURE DETECTOR

Abstract. Vision-aided inertial navigation is a navigation method which combines inertial navigation with computer vision techniques. It can provide a six degrees of freedom navigation solution from passive measurements without external referencing (e.g. GPS). Thus, it can operate in unknown environments without any prior knowledge. Such a system, called IPS (Integrated Positioning System) is developed by the German Aerospace Center (DLR). For optical navigation applications, a reliable and efficient feature detector is a crucial component. With the publication of AGAST, a new feature detector has been presented, which is faster than other feature detectors. To apply AGAST to optical navigation applications, we propose several methods to improve its performance. Based on a new non-maximum suppression algorithm, automatic threshold adaption algorithm in combination with an image split method, the optimized AGAST provides higher reliability and efficiency than the original implementation using the Kanade Lucas Tomasi (KLT) feature detector. Finally, we compare the performance of the optimized AGAST with the KLT feature detector in the context of IPS. The presented approach is tested using real data from typical indoor scenes, evaluated on the accuracy of the navigation solution. The comparison demonstrates a significant performance improvement achieved by the optimized AGAST.