Chapters
00:03
Intro: Photographing Grocery Products with Text
00:17
Photography to Publishing in Under 1 Minute
00:31
Example Angles of View
00:42
Image Capture & Download
00:54
Optical Character Text Recognition
01:18
Extracting Text with Unwrap Tool
01:36
Adding Labels to Structure Text
01:45
Delivery of Extracted Text
01:55
Discover More Tools Like OCR Today
Summary
Follow along this video demonstration of PhotoRobot Controls App features for Optical Character Recognition (OCR) and GS1-compliant image capture. The demo shows how PhotoRobot’s OCR is able to read and extract text on packaging and labels wrapping around items. Watch as we show how to produce all necessary product imagery while also automatically recording product information. This includes weights, dimensions, ingredients, nutritional and on-package data. We show the workflow of capturing and publishing a number of grocery products and canned goods – all in under 1 minute. Find out how we produce 360 spins in tandem with GS1 images, from packaging, to informational, planograms, and marketing photos. The OCR feature with the Unwrap tool makes it easy to read, extract, and edit text on labels. Combine that with PhotoRobot automatic delivery, so all processes become “set once and forget” with PhotoRot automation.
Transcript
00:04 When photographing grocery products, we often work with objects like cans and jars, which are rounded. I will show you that PhotoRobot can not only photograph these super-fast, but can also read the text on the packaging – in spite of the fact that the text wraps around the object.
00:17 Let's show you how. As always, we're going to play our favorite game - each product online in under one minute. In that time, PhotoRobot spins the object, places the camera at the centre of the product's height, as well as the selected angle of view.
00:32 For example, 0° looks almost like rolling it on a flatbed scanner... Or, 15 degrees is very useful as a marketing image. We can select any angle up to 90°. PhotoRobot doesn't pause the rotation to take each image. It just freezes the movement at the right time. Images are then automatically downloaded to the computer. Or, even better, directly to the cloud for automated processing.
00:54 Text recognition can be very useful. For example, when you want to get the nutrition facts on your product web page such as these. Or maybe you want to validate the information that you receive from your suppliers. And that's where OCR comes in very handy. Let's test it.
01:09 We need to take into consideration that the smaller the diameter of the can, the more distortion there is in the text. This makes it nearly impossible to read. Luckily, we have a special tool called "unwrap", so let's go ahead and use it.
01:21 As you can see, it makes the can itself distorted, but the lines of text are now perfectly straight. This can be used as a separate image, or we can easily apply our OCR feature by clicking this little icon.
01:34 And just like that, we get the full transcription of the text. Now, we can add labels. For example, let's give this part here a name. This helps us structure the text and decide which parts we want to keep.
01:45 The extracted text can be exported in JSON or XML feeds, the same ones that you are using for the automated delivery of your image content. So it's "set and forget".
01:55 And this is just one of the many special tools that come with your license of PhotoRobot_Controls. Our OCR tool can recognise about 100 different languages, probably including yours. So why not get in touch with us and let's discuss what we can do for you. Thank you for watching.
Watch Next
View a production demo of on-mannequin photography using the PhotoRobot Cube and Controls App Workflow Automation Software.
Discover the design and dynamics of PhotoRobot’s Centerless Table for automated 2D + 360 + 3D photography in this production video demo.
