Waiting Line Estimation Using Congestion Correction Coefficient
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for estimating waiting line waiting times suffer from reduced accuracy and significant errors, particularly when the waiting line is congested and multiple people overlap in the captured image, leading to inaccuracies in person identification and distance calculation.
Innovation Solution
An information processing apparatus that detects objects in a waiting line, corrects the object count using a correction coefficient when it exceeds a predetermined threshold, and estimates waiting time based on the corrected count and the number of objects leaving the line per unit time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If image analysis is used to detect objects in a waiting line, then automated waiting time estimation is achieved, but detection precision deteriorates when multiple people overlap in congested conditions
Solution Approach 1:
The patent introduces a correction coefficient as an intermediary factor to mediate between the automated detection system and the actual person count. When congestion is detected (number of detected objects exceeds threshold), the correction coefficient adjusts the detected count to compensate for overlapping objects that reduce identification accuracy, thereby maintaining measurement precision while preserving automation.
Solution Approach 2:
The patent changes the parameter used for waiting time estimation from the raw detected object count to a corrected count that incorporates the correction coefficient. This parameter transformation accounts for detection errors in congested conditions, allowing the system to maintain high measurement precision without sacrificing automation extent.
2Device complexity
If conventional image analysis is used in congested waiting lines, then system complexity remains low, but waiting time estimation accuracy deteriorates
Solution Approach 1:
The correction coefficient serves as a simple intermediary mechanism that significantly improves waiting time estimation accuracy without adding substantial system complexity. It requires only threshold comparison and multiplication operations, maintaining computational simplicity while effectively compensating for detection errors in congested conditions.
Solution Approach 2:
The patent transforms the raw detection parameter into a corrected parameter through a simple mathematical operation (multiplication by correction coefficient). This parameter change dramatically improves estimation accuracy while adding minimal computational complexity, resolving the contradiction between simplicity and precision.
3Speed
If object detection is performed without correction in congested conditions, then calculation speed remains high, but traveling distance calculation accuracy deteriorates
Solution Approach 1:
The correction coefficient acts as a computationally efficient intermediary that quickly adjusts detected object counts to account for congestion. The operation involves simple threshold comparison and multiplication, maintaining high calculation speed while significantly improving traveling distance and waiting time estimation accuracy.
Data Source
AI summary
An information processing apparatus estimates a waiting time in a waiting line configured by a plurality of objects and includes at least one processor coupled to a memory, serving as a first obtaining unit to obtain a number of objects detected in an image in which the waiting line is captured, a second obtaining unit to obtain a number of objects that leave the waiting line per unit time, a correcting unit to correct the first obtained number by multiplying a correction coefficient with the first obtained number, in a case when the first obtained number is equal to or greater than a predetermined threshold, and an estimation unit to estimate the waiting time in the waiting line based on the corrected number and the second obtained number, when the first obtained number is equal to or greater than the predetermined threshold.