River Condition Prediction Using Image and Topography Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting river conditions, such as those described in Patent Literature 1, are limited in accuracy when predicting water levels at points away from installed measurement devices or where the number of measurement devices is insufficient.
Innovation Solution
A prediction method that utilizes river images captured at specific positions, along with associated capturing position information, to detect river condition information. This information is then combined with topography information to predict river conditions at different points along the river.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If measurement devices are installed at all points along the river to improve prediction accuracy, then the accuracy of predicting river conditions improves, but the device complexity and cost increase significantly
Solution Approach 1:
The patent uses images captured by cameras as copies or proxies for direct measurement device readings. Instead of installing measurement devices at every point, the system captures visual images of the river and uses image processing to extract water level and flow information, effectively replacing physical measurement devices with optical copies that can be obtained more easily and at lower cost
Solution Approach 2:
The patent introduces an intermediary processing system that includes image processing units and prediction units. These intermediaries translate images captured at certain locations into predictive information about river conditions at other locations, using topography information and hydrological models as mediating elements to bridge the gap between measurement points and prediction targets
2Device complexity
If measurement devices are installed only at limited points to reduce device complexity, then the device complexity decreases, but the prediction accuracy at points away from measurement devices deteriorates
Solution Approach 1:
The patent transitions from a one-dimensional approach (measurement devices at discrete points along the river) to a two-dimensional approach by incorporating topography information and spatial relationships. The system uses the vertical dimension (topography, elevation) and horizontal spatial relationships to predict conditions at any point along the river based on measurements from fewer locations, effectively adding dimensional context to improve prediction coverage
Solution Approach 2:
The patent creates a universal prediction system that can predict river conditions at any location along the river using a single integrated platform. The prediction unit is designed to handle multiple prediction targets simultaneously by incorporating topography information and hydrological models that work across the entire river system, making the measurement system universally applicable rather than location-specific
Data Source
AI summary
A prediction device 100 of the present invention includes a detection means 121 for, on the basis of a river image that is an image obtained by capturing a river and associated with capturing position information representing a position where the river is captured, detecting river condition information representing a condition of the river at the position where the river image is captured; and a prediction means 122 for, on the basis of the capturing position information, the river condition information, and topography information representing the topography of the river, predicting a river condition representing a condition of the river at a given point of the river. The given point is different from the position represented by the capturing position information.