Commodity Pickup Apparatus Using Multi-Angle Image Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing commodity pickup systems in stores struggle to accurately identify and differentiate between heavy and lightweight items, such as a 10 kg rice bag and items of several tens of grams, using weight sensors, leading to difficulties in correctly sorting and recognizing commodities in mixed loads.
Innovation Solution
A commodity pickup apparatus equipped with cameras and illumination LEDs that capture images from multiple angles, perform object recognition, and use image processing to detect and identify commodities based on their presence and type, enabling accurate detection and sorting regardless of the item's weight or orientation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If weight sensors are used to identify commodities, then the sorting process is simple and fast, but it becomes difficult to accurately identify both heavy and lightweight items in mixed loads
Solution Approach 1:
The patent combines multiple detection methods (weight sensing and image recognition) into a unified sorting system. The weight sensor provides initial filtering while the image recognition system provides precise identification, creating a complementary multi-modal detection approach that resolves the contradiction between speed and accuracy.
Solution Approach 2:
The image recognition system acts as an intermediary between the weight sensor and the sorting mechanism. When weight-based identification is ambiguous or insufficient, the system introduces visual verification through cameras and image processing algorithms to accurately distinguish between different commodities regardless of weight differences.
2Device complexity
If a single weight sensor is used for all commodities, then the device complexity is low, but the reliability of commodity identification decreases when handling mixed weights
Solution Approach 1:
The system employs a multi-functional detection approach where the same imaging system serves multiple purposes: identifying commodity types, verifying weight-based classifications, and providing visual records for sorting verification. This universal imaging component enhances reliability without proportionally increasing overall system complexity.
Solution Approach 2:
The patent replaces purely mechanical weight-based identification with an optical recognition system that uses image processing to identify commodities. This substitution enables reliable detection of both heavy and lightweight items using the same detection mechanism, improving reliability while maintaining reasonable system complexity through software-based discrimination.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system ensures accurate identification and sorting of commodities by using image processing and object recognition, allowing for efficient and reliable pickup of ordered items, even when mixed with other items of varying weights.
Implementation Method 1
The controller detects a fact that the commodity picked up is put in the accommodation portion and commodity information of the commodity put in the accommodation portion, on the basis of an image obtained by imaging the accommodation portion by the imaging device
Data Source
AI summary
According to an embodiment, the controller of a commodity pickup apparatus obtains information related to a commodity to be picked up and the number of pieces thereof. Further, the controller detects a fact that a commodity picked up is put in the accommodation portion and commodity information of the commodity put in the accommodation portion, on the basis of an image obtained by imaging the accommodation portion. The display device of the commodity pickup apparatus identifies the commodity to be picked up as a commodity picked up, a commodity to be picked up next, or a commodity not yet picked up and displays the commodity.