A liquid level detection method for transparent bottles based on machine vision

Abstract

The invention discloses a method for detecting liquid level in transparent bottles based on machine vision, which includes the following steps: 1. Selecting a plurality of standard liquid level images and converting them into grayscale images; 2. Obtaining each standard liquid level through a template matching method The reference height value of the grayscale image; 3. Obtain the actual height value of each standard liquid level grayscale image by one-dimensional measurement; 4. Calculate the average value of the reference height value obtained in step 2, and calculate the average value of the reference height value obtained in step 3. Calculate the average value of the actual height value and calculate the standard height value; 5. Obtain the image of the liquid level to be measured and convert it into a grayscale image; 6. Obtain the actual height value of the grayscale image of the liquid level to be measured by one-dimensional measurement, and calculate the value of the liquid level to be measured The height error between the liquid image and the standard liquid level image; if the height error is less than or equal to the tolerable error, it is determined that the liquid level in the liquid level image to be tested is qualified, otherwise it is determined that the liquid level is unqualified. The method disclosed by the invention can accurately measure the height of the liquid level, and judge whether the product is qualified in combination with production requirements.

Description

technical field [0001] The invention belongs to the technical field of machine vision, and in particular relates to a machine vision-based liquid level detection method for transparent bottles. Background technique [0002] Liquid level detection refers to certain methods and technical means to measure the height of liquid. This technology is widely used in liquid food production testing, pharmaceutical manufacturing and testing, liquid chemical testing and other fields. Due to manual misjudgment and detection equipment damage to the body test liquid will seriously affect the quality of the product, especially for pharmaceuticals and dangerous chemicals, the accuracy of the capacity is directly related to the quality of the product. [0003] The current liquid level detection mostly adopts pressure difference method, sensor, manual visual inspection, etc. These detection methods have high requirements on hardware equipment, high equipment price, large mechanical loss, are e...

AI Technical Summary

Problems solved by technology

[0003] The current liquid level detection mostly uses pressure difference method, sensor, manual visual inspection, etc. These detection methods have high requirements on hardware equipment, high equipment price, large mechanical loss, are easily affected by external environmental factors, and require more manual intervention, so the detection speed and accuracy not tall

In the food and drug manufacturing industry, due to the high environmental sanitation requirements, it is difficult to ensure a clean production environment due to too many people on the production site

Especially in the pharmaceutical industry, the accuracy of drug dosage is an extremely important indicator. It is difficult to achieve high-precision standards with low-precision testing equipment and methods.

Images