Glass Bottle Liquid Level Detection Technology: An Innovation from Traditional Mechanical Methods to Intelligent Vision

In industries such as beverages and pharmaceuticals, precise liquid level control is crucial for product quality and corporate reputation.

Accurate detection of liquid levels in glass bottles has become a critical aspect of modern industrial production. Traditional manual visual inspection methods are inefficient and prone to errors, failing to meet the demands of large-scale production. With technological advancements, advanced detection technologies based on machine vision and photoelectric sensing have emerged, significantly improving detection efficiency and accuracy.

These technologies not only overcome challenges such as uneven glass bottle wall thickness and variations in liquid transparency, but also adapt to the real-time detection requirements of high-speed production lines, providing companies with reliable quality control solutions.

1. Technological Evolution: The Shift from Manual to Intelligent Detection

In the industrial production field, the development of glass bottle liquid level detection technology has undergone an evolution from simple to complex, from manual to intelligent. Initially, production companies generally relied on manual visual inspection, where workers observed the liquid level in the glass bottles with their naked eyes to determine product quality.

This method heavily relies on the experience and attention of the personnel; prolonged work can lead to visual fatigue, resulting in misjudgments or missed detections. With the expansion of production scale and increased speed, manual inspection can no longer meet the demands of modern industry for efficiency and quality.

The introduction of machine vision technology has completely changed this situation. Computer vision-based detection systems acquire images of the liquid level in glass bottles using image acquisition equipment, and then analyze the liquid level position using digital image processing algorithms.

These systems can operate continuously and stably without being affected by fatigue, greatly improving detection efficiency and reliability. Currently, machine vision detection methods have become the mainstream technology for liquid level detection and are widely used in the beverage, beer, and pharmaceutical industries.

In recent years, with the development of deep learning technology, liquid level detection has entered a new intelligent stage. The DLIA industrial defect detection system uses deep learning algorithms, training a large amount of image data to enable the system to accurately identify the liquid level height and shape characteristics inside the glass bottle.

This artificial intelligence-based detection method has strong adaptability and anti-interference capabilities, enabling accurate detection even under complex backgrounds or special lighting conditions, representing the future development direction of liquid level detection technology.

2. Working Principle of Machine Vision Detection Methods

A machine vision detection system typically consists of an image acquisition module, an illumination module, an image processing and recognition module, and an output control device. These components work together to achieve precise detection of the liquid level in glass bottles.

During image acquisition, the system uses an LED white backlight panel to illuminate the liquid level position of the glass bottle under test, and the camera captures the liquid level image from the front of the bottle. To obtain a clear liquid level line, the camera and light source parameters need to be finely adjusted so that the liquid level in the glass bottle is clearly separated from the beverage below.

This backlighting method produces high-contrast images, laying the foundation for subsequent image processing.

The image processing stage includes several steps. First, the acquired image is preprocessed using median filtering to remove noise interference and enhance the detection target information.

Next, the preprocessed image is binarized, adjusting the appropriate threshold according to the actual environment to clearly separate the liquid level from the background. A crucial step is to use a geometric matching method to locate the bottle cap position, establishing a coordinate system with the geometric center of the bottle cap as a reference.

Based on the principle that the acceptable liquid level region remains unchanged relative to the geometric center of the bottle cap, the system sets the region of interest (ROI) for the acceptable liquid level region. Finally, the geometric matching method is used again to identify the liquid level line and determine whether it is within the acceptable region.

For different types of liquids, the machine vision system uses different imaging strategies. For colored liquids, a standard backlight is usually sufficient; however, for transparent liquids or situations where the bottle body has character or color interference, a more penetrating infrared backlight is required.

In terms of specific measurement methods, two linear measuring devices can be used to obtain data for the bottle bottom and liquid level, and then calculate the distance between them. The shape matching results are used as a correction source for the linear measuring devices, allowing the measuring devices to follow the product movement and improve measurement accuracy.

3. Detection Solutions for Special Application Scenarios

In some special application scenarios, standard machine vision detection methods may face challenges. For example, in the beer production field, newly filled beer bottles contain a large amount of foam, and traditional methods of measuring through the bottle body are affected by uneven bottle wall thickness.

To address this specific situation, an innovative detection solution has emerged. This solution utilizes the color difference between beer foam and the beer itself, directly capturing an image of the intersection of the liquid and foam using a camera.

The image processor identifies the boundary between the liquid and foam to determine whether the liquid level is acceptable. Because it does not require visible light or infrared light to penetrate the bottle body, this method effectively avoids the influence of uneven bottle wall thickness on the detection results. In the medical field, liquid level detection in infusion bottles has specific requirements. A device based on photoelectric detection technology achieves liquid level monitoring through the collaboration of a laser emission module and a receiving module.

This device positions the laser emission module at the warning line on one side of the infusion bottle, and the laser receiving module on the other side. When the liquid level is above the warning line, the laser beam can pass through the bottle and be detected by the receiving module; once the liquid level drops below the warning line, the direction of light refraction changes, the signal from the receiving module weakens, and the system triggers an alarm.

This non-contact detection method is particularly suitable for the application needs of medical environments.

For situations requiring precise measurement of liquid level height, traditional mechanical measurement methods still have their value. A glass bottle liquid level line measuring tool includes a measuring rod with a limit ring. The limit ring contacts the bottle opening, and the lower end of the measuring rod extends to the standard liquid level line position.

This tool also has a liquid extraction cavity inside. When the liquid level is too high, it can be adjusted to the standard height by extracting liquid. Combined with the measuring scale, it can intuitively determine whether the liquid level is accurate. Although this method is not as efficient as automatic detection, it is suitable for small-batch production or laboratory environments.

4. Comparison of Technical Advantages and Applicable Scenarios

Different liquid level detection technologies have their unique advantages and application scenarios. Machine vision detection methods are non-contact and highly efficient, making them suitable for online detection on high-speed production lines.

The algorithms used in these systems are simple, with few steps, enabling rapid image processing. They are less affected by the environment and have high accuracy. In addition, machine vision systems can be added to existing production lines, enabling online detection during production, greatly improving production efficiency.

Photoelectric detection technology excels in specific fields due to its small size, low cost, and high sensitivity. Especially when combined with a microcontroller to process the optical signal, it makes system judgment and control more precise.  It can also implement anti-interference measures through software algorithms to achieve delayed multiple comprehensive judgments, avoiding false positives and false negatives.

This method is particularly suitable for scenarios such as medical infusions, where equipment size and cost are critical.

Image detection technology specifically designed for special applications such as beer bottles solves the problem of uneven bottle wall thickness that traditional methods struggle with. By directly identifying the boundary between the liquid and foam, it avoids errors caused by measurements through the bottle body, greatly improving the reliability of the detection results. This method provides accurate detection results and has a wide range of applications, especially suitable for testing liquids that easily produce foam, such as beer and carbonated beverages.

Although traditional mechanical measurement methods are less efficient, they have the advantages of simple principles and low cost, making them suitable for small-batch production or laboratory environments. Without the need for complex electronic equipment and algorithms, accurate liquid level height can be obtained through intuitive physical contact, retaining its irreplaceable value in specific situations.

5. Future Development Trends and Challenges

With technological advancements and industrial upgrading, glass bottle liquid level detection technology is developing towards greater intelligence and precision. The application of deep learning algorithms has become an important trend. By training a large amount of image data, the system can adapt to various complex scenarios, improving detection accuracy and robustness.

In the future, AI-based detection systems will be able to autonomously learn and optimize algorithms, continuously adapting to new products and environments, reducing the workload of manual debugging and maintenance.

Another significant trend is the integration of multiple technologies. Hybrid detection systems combining the advantages of machine vision, photoelectric sensing, and ultrasound will become a development direction.

These systems can select the most suitable detection method based on different product characteristics, or even use multiple technologies simultaneously for mutual verification, further improving the reliability of detection results. For example, in the highly demanding medical field, photoelectric and vision technologies can be combined to build multiple safety assurance mechanisms.

Real-time performance and high speed are also important directions for the future development of liquid level detection technology. As production line speeds continue to increase, higher demands are placed on the real-time processing capabilities of the detection system.

Future systems will need to adopt more advanced image sensors and more efficient processing algorithms to meet the pace of high-speed production lines. At the same time, the detection system needs to be more closely integrated with other parts of the production line to form a complete intelligent manufacturing ecosystem.

Despite significant technological advancements, liquid level detection still faces some challenges. For example, stability in complex lighting environments, adaptability to bottles of special shapes, and the identification of liquids with extremely low contrast still need further solutions.

In addition, system cost and ease of use are also important factors affecting technology promotion. Future efforts need to focus on continuous innovation in these areas to make liquid level detection technology more mature and widespread.

Through the scenario of a hospital infusion room, we can see the practical application of photoelectric detection technology. When the medication is about to run out, the system automatically sends an alarm, reminding medical staff to take timely action. Behind this seemingly simple technology is the combination of precise photoelectric sensing and intelligent judgment algorithms. With the continuous development of deep learning and sensor technologies, glass bottle liquid level detection will become more accurate and efficient. Future detection systems may integrate multiple sensing technologies, adapting to different bottle types and liquid characteristics, providing more comprehensive quality control solutions for industrial production.

Detection technology is no longer just an auxiliary part of production, but an important force driving industry progress.