Mobile Camera Time-Series Tracking for Changing Crop Growth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing deep learning-based object tracking technologies struggle to effectively monitor and analyze the continuous growth of crops in smart farms due to shape changes and require numerous cameras, leading to high costs.
Innovation Solution
A system and method utilizing a mobile camera in a smart farm to acquire video and location information, generate time series tracking data through a pre-learned video analysis algorithm, and analyze crop growth, with the ability to detect abnormalities and adjust camera paths for further analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If a large number of cameras are used to monitor large quantities of crops in smart farms, then the coverage and monitoring capability are improved, but the cost increases rapidly
Solution Approach 1:
The patent employs mobile cameras that can move along cables or rails to dynamically cover different areas of the smart farm. Instead of using multiple fixed cameras simultaneously, a single mobile camera traverses through different locations, capturing video data of crops across the entire farm area. This dynamic approach replaces the static multi-camera system with a mobile single-camera system, significantly reducing the number of cameras needed while maintaining comprehensive coverage.
2Reliability
If object tracking technology is used to check time axis information of crop growth, then the ability to track crop changes over time is improved, but the technology fails to uniformly apply to plants whose conditions are continuously changing
Solution Approach 1:
The patent applies preliminary action by capturing video data at multiple time points along the crop growth timeline before performing analysis. The mobile camera systematically records crops at different stages, and the server stores these time-series video data. This preliminary data collection enables subsequent analysis of crop growth changes over time, allowing the system to track how crops evolve from one state to another, thereby achieving reliable time series tracking that adapts to continuous shape changes.
3Measurement precision
If deep learning techniques for CNN-based video analysis are applied to recognize crop features, then the accuracy of crop condition recognition is improved, but the system cannot generate time series tracking data for continuous growth monitoring
Solution Approach 1:
The patent implements continuity of useful action by systematically capturing video data of crops at multiple continuous time points throughout the growth period. The mobile camera continuously records crop conditions, and the server stores these sequential video data. This continuous data collection, combined with deep learning analysis, enables both accurate crop condition recognition and generation of time series tracking data, allowing the system to monitor crop growth continuously rather than at isolated moments.
Data Source
Figure 1
Figure 2
AI summary
Provided is a method of generating time series tracking data for crop growth management. The method includes: acquiring video and location information according to movement of a mobile camera installed in a smart farm; generating video analysis information on a crop from the video based on a pre-learned video analysis algorithm; sorting the video analysis information and the location information based on time information at which the video is captured to generate the time series tracking data; storing the generated time series tracking data; and analyzing and predicting the crop growth based on the time series tracking data.