fbpx

Embedding Computer Vision to Improve Construction Productivity

August 20, 2024 | Blog, News

In the construction industry, a shortage of experienced workers is causing significant delays in implementing construction projects and maintenance tasks. To overcome this challenge, the industry urgently needs tools that can enhance efficiency and productivity.

A prime example is the seismic reinforcement process for bridges, which involves attaching metallic brackets to enhance the structure’s strength and stability during seismic activity. To attach these brackets, holes must be drilled into the concrete bridge. However, the presence of steel wires within the concrete often prevents the creation of holes in the planned locations. As a result, the exact positions of the holes are measured after drilling. Traditionally, the positions of the holes are measured using film and pens, a laborious method that is prone to errors.

There are three major challenges with this approach:

  1. It is time-consuming 
  2. Mistakes are inevitable
  3. The process varies between workers, leading to inconsistencies

To address these challenges, we at Top Data Science have developed a solution that automates the measurement process using smartphone camera images. With this approach, you can obtain the measurement results simply by taking pictures with your smartphone. The results are calculated in seconds, drastically reducing time and minimizing errors.

Here are some key technical components to achieve high-precision measurement:

  1. Camera calibration: Each device has a slightly different lens profile, which can introduce errors to the measurement results. Since customers need measurements with millimeter precision, we built a calibration function into the application. This function automatically measures the lens parameters by taking multiple images of objects of known size. Since this is a one-time calibration, users do not need to recalibrate after the initial setup.
  2. Perspective correction: To convert pixels to physical size, we attach tags to the corners of the target object. This allows us to measure the size accurately from the image. Additionally, using tags enables perspective correction. Construction sites often present obstacles that prevent users from taking photos directly in front of the target object. By applying perspective correction, we can adjust the image to appear as if taken from the front, ensuring accurate measurements.

By combining these techniques, we achieved precise measurements of the holes. The following picture shows an example result of a hole measurement.

Raito Kogyo, one of the industry leaders in foundation engineering in Japan, has begun testing this solution in its seismic reinforcement process this spring. This partnership aims to streamline their operations, enhance accuracy, and ultimately improve productivity in their projects.

Author:
Shun Hirai, CEO

AISA on the Home Stretch

AISA on the Home Stretch

AISA (AI based Situational Awareness) is a Business Finland funded co-innovation project belonging to a Nokia-led Veturi program. The technical focus of AISA is on artificial intelligence and sensing technologies, especially video, image and audio processing. The...

What you may love to read