3D Volume Determination Using Depth Image Super Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for determining the volume of objects moved by industrial trucks are inaccurate, costly, and disrupt logistics processes, as they require stationary measurement and complex setups, especially when dealing with irregular objects.
Innovation Solution
A device using two depth image sensors to create a 3-D model of the object by super-resolving low-resolution sensor data, allowing for accurate volume determination without interrupting logistics processes, and employing truncated signed distance functions to correct measurement errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If strip-based laser scanning is used to sense the contour of loaded goods, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces expensive, complex strip-based laser scanners with inexpensive depth image sensors (such as Kinect cameras) that capture depth information in the form of images. These sensors are significantly cheaper and simpler while providing sufficient measurement precision for volume determination of loaded goods.
Solution Approach 2:
The patent substitutes the mechanical strip-based laser scanning system with an optical imaging-based depth sensing system. Instead of using laser lines and mechanical scanning, the system uses depth image sensors that capture the entire field of view simultaneously, eliminating the need for complex mechanical scanning components.
2Measurement precision
If stationary measurement is used to determine object volume, then measurement precision is improved, but productivity decreases due to workflow interruptions
Solution Approach 1:
The patent transitions from stationary measurement to dynamic measurement by capturing depth images of moving loaded goods at multiple positions along the conveyor belt. The system processes these sequential images to reconstruct the 3D shape and calculate volume, enabling measurement without stopping the logistics workflow.
Solution Approach 2:
The patent performs preliminary capture of depth images at multiple positions before final processing. By capturing images at several positions along the conveyor belt in advance and then processing them together, the system enables volume determination without interrupting the continuous logistics operation.
3Measurement precision
If high-resolution depth image sensors are used, then measurement precision is improved, but device cost increases
Solution Approach 1:
The patent merges multiple low-resolution depth images captured at different positions along the conveyor belt into a single high-resolution 3D representation. By combining information from multiple lower-resolution images taken from different viewpoints, the system achieves the measurement precision of a high-resolution sensor without using one.
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
Enables reliable, cost-effective, and non-disruptive volume determination of moving objects, improving accuracy and reducing errors by producing a high-resolution 3-D model from low-resolution sensor data, while avoiding the complexities of traditional methods.
Implementation Method 1
a sensor device (3, 4) for purposes of sensing the shape of a surface of an object (1)
Data Source
AI summary
The invention relates to a device for determining the volume of an object moved by an industrial truck and to a corresponding method. The device comprises a first depth image sensor (3) and a second depth image sensor (4), which are arranged in such a way that the object (1) can be sensed from two different directions as the object passes through the passage region (5), wherein the device (10) is designed to produce a sequence of individual images in a first resolution by means of each of the depth image sensors (3, 4), whereby the industrial truck (2) and the object (1) are sensed from different viewing angles as the industrial truck and the object pass through the passage region (5). The device also comprises a volume-determining apparatus (11), which is designed to subject the sequences of the individual images in the first resolution of the first depth image sensor (3) and of the second depth image sensor (4) to a super resolution processing in order to create a 3-D model (17) comprising the industrial truck (2) and the object (1) in a second resolution that is higher than the first resolution and to determine the volume of the object (1) on the basis of the produced 3-D model (17).