| Three-stage process for inspecting structural elements in construction site, involves utilizing LLava-based vision-language model deployed on NVIDIA Jetson platform, and artificial intelligence (AI) model estimates rod diameters with high precision and ensures rapid inference with minimal latency | |
| 2024-03-13 | |
| 专利权人 | SONA COLLEGE TECHNOLOGY (SONA-Non-standard) |
| 申请日期 | 2024-03-13 |
| 专利号 | IN202341084943-A |
| 成果简介 | NOVELTY - The process involves collecting site images, and video images containing reinforcement details. The collected images are processed using artificial intelligence-based vision algorithms, including image preprocessing techniques to extract reinforcement attributes such as bar diameter, spacing, and alignment. The extracted details against structural drawings are verified and discrepancies are identified. A LLava-based vision-language model deployed on an NVIDIA Jetson platform is utilized for real-time reinforcement measurement. The AI model estimates rod diameters with high precision and ensures rapid inference with minimal latency. USE - Three-stage process for virtually inspecting structural elements in construction site such as reinforced concrete columns, reinforced concrete slabs, and reinforced concrete beams. ADVANTAGE - The process enables instant and reliable reinforcement analysis directly at the construction site, eliminating the need for time-consuming manual inspections. The process enhances construction site efficiency, minimizes manual workload, and ensures consistent and accurate reinforcement inspections, ultimately contributing to the improved execution and safety of construction projects. By leveraging real-time processing capabilities, the process ensures that measurements are instantly available to engineers, improving decision-making and reducing project delays. The AI model is fine-tuned using a custom dataset of construction rods to enhance prediction accuracy. DESCRIPTION OF DRAWING(S) - The drawing shows a schematic view of the different diameter of construction rods. |
| IPC 分类号 | G06N-020/00 ; G06N-003/08 ; G06Q-050/08 ; G06T-007/00 |
| 国家 | 印度 |
| 专业领域 | 信息技术 |
| 语种 | 英语 |
| 成果类型 | 专利 |
| 文献类型 | 科技成果 |
| 条目标识符 | http://119.78.100.226:8889/handle/3KE4DYBR/17743 |
| 专题 | 中国科学院新疆生态与地理研究所 |
| 作者单位 | SONA COLLEGE TECHNOLOGY (SONA-Non-standard) |
| 推荐引用方式 GB/T 7714 | KUMAR A P,NAVEENKUMAR A,M D P,et al. Three-stage process for inspecting structural elements in construction site, involves utilizing LLava-based vision-language model deployed on NVIDIA Jetson platform, and artificial intelligence (AI) model estimates rod diameters with high precision and ensures rapid inference with minimal latency. IN202341084943-A[P]. 2024. |
| 条目包含的文件 | 条目无相关文件。 | |||||
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。
修改评论