Machine vision inspection: Liquid bubble size detection solution

In industrial production and experimental testing, the state of air bubbles often directly affects product quality and process stability. For example, in liquid handling, chemical reactions, and related manufacturing processes, the size, distribution, and changes in air bubbles are key testing indicators. Recently, Zhixiang Shijue received a client's request to "detect the size of air bubbles generated in a water bucket," and completed a machine vision-based inspection solution, achieving excellent application results.

Project Background: Significant Challenges in Bubble Detection

In this project, the client wanted to detect bubbles generated after gas was introduced into a water bucket in real time and accurately analyze their size. However, the following technical challenges arose during the actual detection process:

Unstable bubble morphology: Bubbles are constantly generated, rise, and burst in water, resulting in rapid morphological changes.

Significant interference from transparent media: Both the water and the bubbles are transparent or semi-transparent, leading to low image contrast.

Obvious effects of light refraction and reflection: Easily causing blurred edges or misjudgments.

High requirements for dynamic detection: Recognition and measurement need to be completed quickly during movement.

These factors place high demands on the imaging quality and algorithm processing capabilities of the vision system.

Solution: Software and hardware co-optimization

By analyzing the aforementioned challenges, the technical team planned to address both hardware selection and algorithm optimization, constructing a stable and efficient visual inspection solution:

Imaging System Optimization: A high-resolution industrial camera, combined with a suitable focal length lens, was selected to ensure clear visibility of bubble details.

Light Source Design: Backlighting or structured light illumination optimized for transparent media was employed to enhance bubble contour contrast.

Image Processing Algorithms: Automatic bubble recognition and measurement were achieved through edge extraction, contour analysis, and size calculation algorithms.

Dynamic Acquisition Capability: High-speed image acquisition was supported to meet the detection requirements of real-time bubble changes.

Through systematic debugging and parameter optimization, the boundaries of bubbles in the image were made clearer, effectively improving detection stability.

Testing results: Dimensions are clearly measurable, and results are stable and reliable.

In actual testing, the vision system was able to stably capture bubbles in water and clearly present their outlines. Through algorithmic processing, it achieved accurate measurement of bubble diameter and differentiated and statistically analyzed bubbles of different sizes.

Overall detection performance was as follows:

Clear bubble edge recognition and complete outlines;

Stable size measurement results with good repeatability;

Adaptable to dynamic changes in bubbles, enabling continuous detection;

Effectively reduced human intervention and improved detection efficiency.

This solution has met the customer's core needs for bubble size detection, providing a reliable technical foundation for subsequent application implementation.

This bubble detection project not only validated the feasibility of machine vision in transparent media inspection but also further enhanced the application capabilities of vision technology in fluid and dynamic scenarios. Compared to traditional manual observation methods, visual inspection offers higher efficiency and consistency, providing quantitative data support for industrial process control.

As industrial inspection needs become increasingly refined and sophisticated, machine vision is gradually extending from traditional appearance inspection to more complex dynamic and transparent media scenarios. Zhixiang Shijue will continue to optimize its visual inspection solutions based on practical application needs, providing customers with more accurate and stable industrial vision solutions to help enterprises achieve intelligent upgrades and high-quality development.