In this article, we will discuss the app features that help take photos of shelves with the quality needed to be properly processed by AI algorithms.
The Eyrene mobile application is a powerful tool for merchandisers who visit stores and execute shelf audits.
Merchandisers are busy as bees
Merchandisers are usually quite busy with tasks that they must execute during short in-store visits. The ordinary task list may contain checking shelf tags with sales price, ensuring that displays are stocked with the correct products, and confirming that the products are placed on the proper shelves.
What’s more, for brands of any size, it’s pretty expensive to hire in-store merchandisers, so CPG managers strive to assign as many as possible tasks to merchandisers.
Merchandisers have quite a tough job, requiring patience, attentiveness, and the ability to act fast in any situation. The number of people ready to work this way for long periods of time is not great, as lots of merchandisers prefer to move on to some other position that’s less stressful. As a result, experienced in-store merchandisers are valuable. The rest are mostly new to the merchandising routine and have to adapt on the go. Therefore, novices need mobile apps providing merchandisers with all possible means to automate their routine and deliver the needed quality.
How the Eyrene mobile app helps
The Eyrene mobile app has lots of various features, but now we will discuss the features that help merchandisers take photos of the shelves which are suitable for further analysis by neural networks.
Image quality is very critical for image recognition accuracy. The quality of data processing depends on the quality of the raw data. That means the images should be taken in conditions (enough light in the store, good angle) most suitable for further image processing and extracting important data.
With the help of the Eyrene app, three core issues can be resolved:
- Merchandisers can take photos with quality good enough for valuable data extraction by algorithms.
- Specific features in the solution are designed to monitor the image quality.
- If an image is taken with low quality, the user gets an alert and guidelines on how to improve it.
Eyrene features that simplify taking photos of shelves
In the Eyrene mobile app, there are a number of features to help merchandisers take photos of the shelves, even if they are not sure how to do it.
This effect is achieved with the help of two stripes in the middle of the screen, one representing the level, and the other representing the horizon. The merchandiser should have these stripes aligned. Once that happens, they turn into one yellow stripe, which means that the view of the shelf is good enough to take an image.
This feature helps to make certain adjustments if needed. However, situations may occur when it’s almost impossible to take a proper photo, but algorithms in the app still make it possible to improve some of the most critical issues. Deviations in the angle usually affect the accuracy of image recognition, and for the algorithms, it becomes a challenging task to identify packaging size. Any deviations in the angle of view might lead to a decrease in image recognition accuracy.
Image recognition accuracy is one of the factors which influence the points merchandisers to receive for tasks accomplished during their in-store visit. The higher the accuracy of image recognition, the more products are identified in the photo, and the greater financial compensation the merchandiser receives.
#2 - Helps to take photos of the goods stored in horizontal refrigerators (for example, frozen goods).
The key difference here is that the phone should be rotated into a horizontal position. To use this feature properly, the user must align two dashes, one of which is white and is permanently located right in the center of the screen, and the other of which is yellow and dynamically moves around the screen when the user moves the phone.
Both features (for vertical and horizontal image capturing) are activated automatically depending on the position in which the user holds the phone.
#3 - Helps merchandisers take panoramic photos.
This is a really good tool for taking panoramic photos of large or long shelves in hypermarkets, as well as of the shelves located in a smaller convenience stores, where they are placed quite close to each other.
This tool sticks several images of the shelf together into one panoramic photo, which is subsequently analyzed to identify the shelf itself, the goods located on it, price tags, etc.
Even though any modern smartphone has this feature to take panoramic photos, it is not adapted for the requirements of image recognition algorithms, as it allows the user to take photos only in one direction.
For the purposes of in-store panorama, the feature should work the other way. Commonly, the shelf is so large that it can’t be squeezed into the photo, either vertically or horizontally. Thus, the panorama is taken in several steps while the phone is moved in a zigzag manner. One more important thing is that the software prompts the user to take photos with overlap. The task for the user is just to follow the instructions provided by the app on the screen.
