Registration number: KK-220021-A
Technical name: Flow velocity and flow rate measurement system [Hydro-STIV]
https://www.netis.mlit.go.jp/netis/pubsearch/details?regNo=KK-220021%20

STIV Overview

1.Theory: Flow velocity measurement using Space-Time Image

Left: Generating a Space-Time Image (STI) from video

The yellow lines shows search lines, and the red squares show features of brightness value(surface ripples etc). STI is generated by arranging the brightness values on the search line in the time direction, and the stripe pattern indicates velocity.

Right: Calculation of flow velocity (V) from Space-Time Image (STI)

Velocity is calculated using slope of the stripe pattern(Φ), length, and time of STI. When Φ is small, velocity is slow(left-bottom) and vice versa(right-bottom).

2.Evaluation of STIV measurement accuracy

3.AI enables high-speed and high accuracy measurement

• Resistant to noise such as water droplets and blurring, automating image analysis
• Enables automatic and stable flow velocity and discharge measurement in various environments such as adverse weather conditions

AI (deep learning) enables more stable measurement than conventional methods. Highly accurate measurement can be performed with complete automatic analysis even when obstacles are in the image, such as when there is muddy water, heavy rain with high water, and there is a reflecting light at night and an infrared camera is used.

4.Feature

Point1：Drone footage for measurement

Images taken from the sky by a drone or other devices can be used to measure the flow velocity and discharge. Wide rivers, mountainous areas, or rivers during flood can be easily measured by drone footage. Another advantage of vertical shooting is that no geometric correction is required since the actual scale can be detected by the distance between only 2 points in the image.

Point2：Nighttime measurement using an ultra-high-sensitivity camera

Example of video taken at 9:30 p.m.

Normal camera

Ultra-high-sensitivity camera

By using an ultra-high-sensitivity camera that can take clear images even at night, Hydro-STIV can measure flow velocity and discharge 24 hours a day, even in environments with limited light sources.

Point3：Automatic water level measurement by recognizing the water’s edge on a plane

• Water level measurement function by image
• Automatic water level detection from images without using a water level meter

The water level is measured by detecting the water's edge on vertical structures such as water markers, bridge piers, and concrete walls in the image.
By using a staff gauge or surveying a vertical structure in the image, the water level can be measured on Hydro-STIV software.

Case Study

Othea River @New Zealand

Rangitata River @New Zealand

Tekapo River @New Zealand

Herbert River @Australia

@Melbourne, Victoria, Australia

Russel River @Australia

North River @Canada

Ciujung River @Indonesia

Data