In recent years, object detection algorithms have achieved great success in the field of machine vision. To pursue the detection accuracy of the model, the scale of the network is constantly increasing, which leads to...
详细信息
In recent years, object detection algorithms have achieved great success in the field of machine vision. To pursue the detection accuracy of the model, the scale of the network is constantly increasing, which leads to the continuous increase in computational cost and a large requirement for memory. The larger network scale allows their execution to take a longer time, facing the balance between the detection accuracy and the speed of execution. Therefore, the developed algorithm is not suitable for real-time applications. To improve the detection performance of small targets, we propose a new method, the real-time object detection algorithm based on transfer learning. Based on the baseline Yolov3 model, pruning is done to reduce the scale of the model, and then migration learning is used to ensure the detection accuracy of the model. The object detection method using transfer learning achieves a good balance between detection accuracy and inference speed and is more conducive to the real-timeprocessing of images. Through the evaluation of the dataset voc2007 + 2012, the experimental results show that the parameters of the Yolov3-Pruning(transfer): model are reduced by 3X compared with the baseline Yolov3 model, and the detection accuracy is improved, realizes real-timeprocessing, and improves the detection accuracy.
In this paper, the 3D space imaging model of machine vision is constructed. Starting from the traditional machine vision imageprocessing algorithm flow, the image denoising process and target tracking process are opt...
详细信息
In recent years, with the rapid advancement of digital imageprocessing technology, image dehazing has become a focal point in both scientific research and practical applications. This paper presents a study on a deha...
详细信息
It is well known that video images captured by in-vehicle cameras and surveillance cameras are degraded due to noise introduced by their shooting environments. In this paper, we remove such noise using a learning meth...
详细信息
Dispersion effects are important in rendering scenes that contain translucent objects. Existing ray-tracing algorithms must rely on high-end GPU hardware to achieve real-time rendering when rendering dispersion. Based...
详细信息
real-time identification of road damage, particularly potholes, is essential for enhancing road safety and minimizing vehicle damage. Traditional road inspection methods are often labor-intensive and inefficient. To o...
详细信息
ISBN:
(纸本)9783031837920;9783031837937
real-time identification of road damage, particularly potholes, is essential for enhancing road safety and minimizing vehicle damage. Traditional road inspection methods are often labor-intensive and inefficient. To overcome these limitations, we propose an automated system that leverages deep learning to detect potholes in real-time from video input. This system can detect potholes under varying lighting conditions, specifically the YOLO (You Only Look Once) model, utilizing advanced computer vision techniques, including daylight, darkness, and vehicle headlights. The system aims to improve road monitoring by automatically identifying potholes and notifying authorities of their locations. By integrating geolocation services, the system pinpoints pothole locations using latitude and longitude coordinates and sends real-time alerts via a Telegram bot. Additionally, image enhancement techniques are employed to optimize detection performance in low-light conditions. The system has demonstrated high accuracy in detecting large potholes and identifying multiple potholes within a single frame. Through automated pothole detection and location sharing, this solution has the potential to significantly enhance road maintenance efficiency, thereby improving road safety and reducing accidents caused by road damage.
Maritime Authorities face significant challenges in monitoring vast maritime domains and enforcing regulations within their Exclusive Economic Zones (EEZs). Synthetic Aperture Radar (SAR) imaging of Fers an effective ...
详细信息
This paper introduces a parallel architecture designed for real-timeimageprocessing applications, utilizing a combination of digital signal processor (DSP) and field-programmable gate array (FPGA) components for opt...
详细信息
This paper addresses the growing interest in deploying deep learning models directly in-sensor. We present "Q-Segment", a quantized real-time segmentation algorithm, and conduct a comprehensive evaluation on...
详细信息
ISBN:
(纸本)9798350383638;9798350383645
This paper addresses the growing interest in deploying deep learning models directly in-sensor. We present "Q-Segment", a quantized real-time segmentation algorithm, and conduct a comprehensive evaluation on a low-power edge vision platform with an in-sensors processor, the Sony IMX500. One of the main goals of the model is to achieve end-to-end image segmentation for vessel-based medical diagnosis. Deployed on the IMX500 platform, Q-Segment achieves ultra-low inference time in-sensor only 0.23 ms and power consumption of only 72mW. We compare the proposed network with state-of-the-art models, both float and quantized, demonstrating that the proposed solution outperforms existing networks on various platforms in computing efficiency, e.g., by a factor of 75x compared to ERFNet. The network employs an encoder-decoder structure with skip connections, and results in a binary accuracy of 97.25 % and an Area Under the Receiver Operating Characteristic Curve (AUC) of 96.97 % on the CHASE dataset. We also present a comparison of the IMX500 processing core with the Sony Spresense, a low-power multi-core ARM Cortex-M microcontroller, and a single-core ARM Cortex-M4 showing that it can achieve in-sensor processing with end-to-end low latency (17 ms) and power consumption (254mW). This research contributes valuable insights into edge-based image segmentation, laying the foundation for efficient algorithms tailored to low-power environments. [GRAPHICS] .
Solid waste management (SWM) is one of the global challenges due to the lack of an efficient automated real-time waste segregation and reuse system. This work deals with the design, development and testing of an intel...
详细信息
暂无评论