Geometric Recognition of Moving Objects in Monocular Rotating Imagery Using Faster R-CNN

Abstract

Moving object detection and tracking from image sequences has been extensively studied in a variety of fields. Nevertheless, observing geometric attributes and identifying the detected objects for further investigation of moving behavior has drawn less attention. The focus of this study is to determine moving trajectories, object heights, and object recognition using a monocular camera configuration. This paper presents a scheme to conduct moving object recognition with three-dimensional (3D) observation using faster region-based convolutional neural network (Faster R-CNN) with a stationary and rotating Pan Tilt Zoom (PTZ) camera and close-range photogrammetry. The camera motion effects are first eliminated to detect objects that contain actual movement, and a moving object recognition process is employed to recognize the object classes and to facilitate the estimation of their geometric attributes. Thus, this information can further contribute to the investigation of object moving behavior. To evaluate the effectiveness of the proposed scheme quantitatively, first, an experiment with indoor synthetic configuration is conducted, then, outdoor real-life data are used to verify the feasibility based on recall, precision, and F1 index. The experiments have shown promising results and have verified the effectiveness of the proposed method in both laboratory and real environments. The proposed approach calculates the height and speed estimates of the recognized moving objects, including pedestrians and vehicles, and shows promising results with acceptable errors and application potential through existing PTZ camera images at a very low cost.

Classification of brain tumours in MRI images using convolutional neural network through Cat Swarm Optimization

“…Neurocognitive networks are huge systems of decentralized and interrelated central nervous neural circuits arranged to conduct cognitive functions. The excitation function substitutes numerous nonlinear functions of the neurocognitive machine, and the pooling procedure is additionally simplified by the convolution operation (Chuang et al, 2020). A fully connected layer manages all interactions established through the earlier layers then works as the classifier to categorize the inputs.…”

Forces balance in coordinate system of object’s existence 3D space

“…Currently, research models have been created that determine the features of trajectories, object height and object recognition using different configurations of hardware implementations 5,6 , analyze the obtained 3D information for object recognition, as well as for perception of their behavior in space. The research allows, for example, in real time to estimate the distance of surfaces for wireless localization of objects 7 , as well as the application of object definition in the field of devices 8,9,14 based on analytical models of object existence in a given industry space.…”