Site icon Visionify

Finding Labeling Errors is Now Easy with Automated Visual Inspection

Finding Labeling Errors is Now Easy with Automated Visual Inspection

When you come across a new product in a store you see it has no label. It means there’s nothing indicating price, ingredients list, or company name. It could be anything like a packet of noodles or a shampoo bottle. Do you know why this happens? Due to a lack of proper label checking solutions.

Future market insights, says the global market for label applicators was estimated at $2.6 billion in 2022 and is projected to grow to $3.5 billion by 2028.

source : Future Market Insights

Finding label errors is a challenge for manufacturing companies. Many labels are used in food manufacturing, from variable information laser printed labels to graphics-intensive, pressure-sensitive labels. Even with advancements in labeling technologies, label error still exists on a large scale and is a major concern for manufacturers. In our blog you will know how to find label errors with an automated visual inspection.

All manufacturers know it is vital to detect incorrect labels swiftly and reliably. It is too time-consuming to manually check product labels with a cosmic speed of today’s production lines. Thanks to digital age food manufacturers can use computer vision technology for quality label checks.

How to implement automated visual inspection in your company?

Step 1: Develop Automated Inspection.

A food production system has several different steps that require an operator to examine a product before being sold to customers. These steps are essential because they help ensure that every person who consumes a food product is safe and healthy. The last step when product leaves production belt is checking if a label is correct. Avoid misbranding and incorrect positioning on product labels to ensure your product is safe to eat.

Human inspectors cannot perform a visual inspection on a scale of billions of dollars worth of products. Human eye can see 10 to 12 images per second where a conveyor belt moves with a speed of 12 inches per second. Visual detection becomes challenging to detect variations in label design when product is smaller.

So to overcome time-consuming manual inspection companies are adopting computer vision automated visual inspection. To process visual information at a much higher speed than a manual operator. In light of this inspection, automation seems to be a future wave.

Automated visual inspection systems usually include:

1. Light source.

2. Cameras.

3. Processing and sorting equipment.

4. Software that analyzes images makes decisions on product quality and takes action accordingly.

Step 2: Image analysis software performs inspection in following ways:

1. Preprocessing images (enhancing edges, removing noise, etc.)

2. Segmenting images (dividing an image into some recognizable sections).

3. Feature extraction.

4. Image classification (analysis of an object can be judged based on a simple pass/fail test.)

Above steps are achieved by putting together a set of machine learning algorithms (convolutional neural networks, support vector machines, and decision trees) and using them to train our system, identify products and ensure consistent labeling. That’s why customized image analysis software 

packages account for differing customer needs.

Step 3: Looking for labeling errors.

Proper labelling issues are among most common problems in food and beverage industry. Automated visual inspection is a great solution. Inspection process includes a variety of tasks such as checking for correct position, scanning characters, and verifying barcodes.

Step 4: Checking position of a label.

Most common error that automated visual inspection looks for is significant incorrect positioning. First label-inspection system checks label to make sure it is where it is supposed to be. Then it inspects label’s positioning making sure that it has been attached to product correctly.

Step 5: Recognizing characters and their meaning.

OCR can help you resolve labeling errors in manufacturing. It examines order of letters in a valuable ingredient list to find mistakes in product descriptions. For instance if we consider product recalls like lemon tarts pack. Ingredients label on product stated presence of cashews but allergen warning statement (“contains”) failed to mention this ingredient is dangerous. So here algorithm learned to link two labels. So it could spot inconsistencies in allergenic ingredients and warn manufacturers before a product is released.

Step 6: Reading barcodes

Barcodes provide information on both products and manufacturers. An incorrect barcode might cause problems for your company if you don’t trace it before shipping. Old labelling systems are not suitable for present production standards.

So adopting a state of art labelling solution is best to skip label errors in your manufacturing facility. You need a camera-based reader to locate and interpret multiple barcodes, regardless of their size and position. And visual inspection systems can benefit from addition of a barcode reader.

How does this solution work for you?

An automated visual inspection system can scan labels on millions of products and flag incorrect or suspicious ones based on data collected from decoding product labels. This information then becomes available for use in further examination or action. For instance it can alert human workers if a product line has a labeling issue. Automated visual inspection systems have become more accurate in recent years and save manufacturers considerable time, effort and money.

What benefits do you get?

1. It adapts quickly to your product needs and surfaces.

2. It’s a customizable solution that can be programmed and monitored remotely.

3. A lot faster than human inspectors. Don’t wait days for an inspection report.

4. Keep your facilities consistent with quality control.

Do you need it?

Rank all sources of label errors in your current production line to determine the ones that affect profit, customer satisfaction, etc., most.

Suppose you want to evaluate whether an automated visual inspection is worth investing in. In that case, you need to consider how much you can gain from identifying and working on specific improvements.

Conclusion

The effectiveness of automated inspection solutions is growing every day. With help of new technology labelling errors can be found with greater ease. Automated inspection solutions identify printing defects present on labels. They can help eliminate manual inspections by humans and give greater accuracy in detection and use in an automated way.

Follow Visionify.ai on LinkedIn & Twitter

Exit mobile version