Objects In Video Images Research Articles

Counting People and Bicycles in Real Time Using YOLO on Jetson Nano

Counting objects in video images has been an active area of computer vision for decades. For precise counting, it is necessary to detect objects and follow them through consecutive frames. Deep neural networks have allowed great improvements in this area. Nonetheless, this task is still a challenge for edge computing, especially when low-power edge AI devices must be used. The present work describes an application where an edge device is used to run a YOLO network and V-IOU tracker to count people and bicycles in real time. A selective frame-downsampling algorithm is used to allow a larger frame rate when necessary while optimizing memory usage and energy consumption. In the experiments, the system was able to detect and count the objects with 18 counting errors in 525 objects and a mean inference time of 112.82 ms per frame. With the selective downsampling algorithm, it was also capable of recovering and reduce memory usage while maintaining its precision.

Detection and Selection of Moving Objects in Video Images Based on Impulse and Recurrent Neural Networks

The purpose of the article is to develop a methodology for automating the detection and selection of moving objects. The detection and separation of moving objects based on impulse and recurrence neural networks simulation. The result of the work is a developed motion detector based on impulse and recurrence neural networks and an automated system developed on the basis of this detector for detecting and separating moving objects and is ready for practical application. The feasibility of integrating the developed motion detector with Emgu CV (OpenCV) image processing package, multimedia framework functions, and DirectShow application programming interface were investigated. The proposed approach and software for the detection and separating of moving objects in video images using neural networks can be integrated into more sophisticated specialized computer-aided video surveillance systems, IoT (Internet of Things), IoV (Internet of Vehicles), etc.

Obtaining Urban Waterlogging Depths from Video Images Using Synthetic Image Data

Reference objects in video images can be used to indicate urban waterlogging depths. The detection of reference objects is the key step to obtain waterlogging depths from video images. Object detection models with convolutional neural networks (CNNs) have been utilized to detect reference objects. These models require a large number of labeled images as the training data to ensure the applicability at a city scale. However, it is hard to collect a sufficient number of urban flooding images containing valuable reference objects, and manually labeling images is time-consuming and expensive. To solve the problem, we present a method to synthesize image data as the training data. Firstly, original images containing reference objects and original images with water surfaces are collected from open data sources, and reference objects and water surfaces are cropped from these original images. Secondly, the reference objects and water surfaces are further enriched via data augmentation techniques to ensure the diversity. Finally, the enriched reference objects and water surfaces are combined to generate a synthetic image dataset with annotations. The synthetic image dataset is further used for training an object detection model with CNN. The waterlogging depths are calculated based on the reference objects detected by the trained model. A real video dataset and an artificial image dataset are used to evaluate the effectiveness of the proposed method. The results show that the detection model trained using the synthetic image dataset can effectively detect reference objects from images, and it can achieve acceptable accuracies of waterlogging depths based on the detected reference objects. The proposed method has the potential to monitor waterlogging depths at a city scale.

Optimization methods of video images processing for mobile object recognition

Recognition of moving objects in video images is mainly based on acquiring the target information in a certain time series. After image processing, relevant algorithms are used to get the internal features and effectively identify the target object. However, image background, noise, definition and other factors will have impacts on mobile object recognition. Therefore, the mobile objects in video images are more complicated than the static objects in the fixed images. The traditional convolutional neural network (CNN) uses gradient descent algorithm for learning and training, and uses gradient descent algorithm to determine the initial thresholds, weights, which may cause the training to fall into a local optimal state. Therefore, this paper proposes an improved adaptive genetic algorithm combined with CNN. The thresholds and weights of CNN can be optimized by using adaptive genetic algorithm (AGA), which can overcome the shortcomings of the original genetic algorithm such as slow convergence. Experimental results shows that the recognition accuracy rate of the experiment increased from 83.75% to 92%, the method can effectively improve the accuracy and efficiency of mobile object recognition.

Potential application of Kanade–Lucas–Tomasi tracker on satellite images for automatic change detection

Monitoring agricultural areas is still a very challenging task. Various models and methodologies have been developed for monitoring the agricultural areas with satellite images, but their practical applicability is limited due to the complexity in processing and dependence on a priori information. Therefore, in this paper, an attempt has been made to investigate the utility of the Kanade–Lucas–Tomasi (KLT) tracker, which is generally useful for tracking objects in video images, for monitoring agricultural areas. The KLT tracker was proposed to deal with the problem of image registration, but the use of the KLT tracker in satellite images for land cover monitoring is rarely reported. Advanced Land Observing Satellite Phased Array type L-band Synthetic Aperture Radar (ALOS PALSAR) data has been used to identify and track the agricultural areas. The tracked pixels were compared with the agriculture pixels obtained from a decision tree algorithm and both results are closely matched. An image differencing change detection technique has been applied after KLT tracker implementation to observe the “change” and “no change” pixels in agricultural areas. It is observed that two kinds of changes are being detected. The areas where agriculture was not there earlier, but now is present, the changes are called positive changes. In the areas where agriculture was present earlier, but now is not present, those changes are referred to as negative changes. Unchanged areas retrieved from both the images are labeled as “no change” pixels. The novelty of the proposed algorithm is that it uses a simplified version of the KLT tracker to efficiently select and track the agriculture features on the basis of their spatial information and does not require a priori information every time.

Segmentation Algorithm Using Method of Edge Detection

This paper attempts to develop a segmentation algorithm applicable to the issue of recognizing objects in video images. The paper presents the steps of the algorithm with a discussion of techniques used in mathematical morphology, filtration and gradient methods. Also there were presented examples of the results of a verification researches.

Multi-feature fusion based fast video flame detection

A video flame detection method based on the multi-feature fusion is presented in this paper. The temporal and spatial characteristics of flames, such as ordinary flame movement and color clues, a flame flickering detection algorithm is incorporated into the scheme to detect fires in color video sequences. An improved Gaussian mixture model method is firstly adopted to extract moving foreground objects from the still background of detection scenes; secondly, detected moving objects are then categorized into candidate and non-candidate flame regions by using a flame color filtering algorithm; finally, a flame flicker identification algorithm based on statistical frequency counting is used to distinguish true flames from fire-like objects in video images. Testing results show that the proposed algorithms are effective, robust and efficient. The processing rate of the flame detection method can achieve 24 fps with image size of 320 × 240 pixels on a PC with an AMD 2.04 GHz processor.

The Relation of Visual Complexity and Objects in Video images

The Relation of Visual Complexity and Objects in Video images