Optical Sensor Adaptive Kalman Filter for Moving Object Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Optical sensors face measurement errors and increased reaction time when detecting moving objects due to position or distance changes, leading to falsified results and lag errors, especially when averaging measured values.

Innovation Solution

The implementation of an adaptive Kalman filter that continuously adjusts filter coefficients to current changes in measured values, allowing for real-time filtering of individual measurements and reducing noise, enabling precise object detection even at high speeds without compromising measurement accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If averaging methods are used to reduce measurement errors, then measurement accuracy is improved, but reaction time is increased and trailing errors occur when objects move relative to the sensor

Engineering Contradiction:
Improvemeasurement accuracyVSAvoidreaction time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies dynamic filtering by switching between different filter types (moving average filter and exponential moving average filter) based on the detected motion state of the object. When motion is detected, the system transitions to a filter that adapts to changing positions, preventing trailing errors while maintaining measurement accuracy. This dynamic adaptation resolves the contradiction by making the filtering process responsive to real-time conditions rather than using a static averaging method.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the filtering parameters dynamically based on motion detection. The system monitors the measured values to detect motion and adjusts the filter coefficients accordingly - using a moving average filter for stationary objects and an exponential moving average filter for moving objects. This parameter change allows the system to maintain both high measurement accuracy and fast reaction time by selecting the appropriate filtering strength based on current conditions.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If averaging methods are used to reduce measurement errors, then measurement accuracy is improved, but trailing errors occur when the object's position changes during averaging

Engineering Contradiction:
Improvemeasurement accuracyVSAvoidmeasurement validity
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The system dynamically switches filtering strategies based on detected motion. When an object is detected to be moving, the patent applies an exponential moving average filter that gives more weight to recent measurements, preventing the trailing errors that occur with standard averaging. This dynamic approach maintains measurement validity by adapting the filtering method to the object's motion state.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent implements feedback by continuously monitoring the measured values to detect motion and adjusting the filter type accordingly. The system uses the measured position or distance values to determine whether to apply a moving average filter (for stationary objects) or an exponential moving average filter (for moving objects). This feedback mechanism ensures that the filtering process maintains measurement validity by responding to real-time changes in the object's position.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If filter coefficients are adjusted continuously to adapt to changing measured values, then measurement accuracy is improved for moving objects, but device complexity increases

Engineering Contradiction:
Improvemeasurement accuracyVSAvoidfiltering system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent changes filtering parameters based on motion detection rather than continuously adjusting all filter coefficients. The system switches between two well-defined filter types (moving average and exponential moving average) based on a simple motion detection threshold. This discrete parameter change approach maintains high measurement accuracy for moving objects while avoiding the complexity of continuously adapting multiple filter parameters simultaneously.

Inventive Principle:
Principle #35Parameter changes

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

This approach improves measurement accuracy and reduces noise in optical sensors, allowing for precise detection of moving objects and faster reaction times by adapting to changing noise power ranges and movement scenarios, thereby enhancing the sensor's performance without increasing costs or structural size.

Implementation Method 1

the optical sensor comprises a transmit/receive unit with at least one transmitter emitting light beams and one receiver receiving light beams. For object detection, the transmitted light beams are typically directed towards the object, reflected by it, and then directed as received light beams to the receiver.

Methodology Applied
Scientific EffectLight reflection: Reflection

Data Source

PatentEP4148460A1Optical sensor and method for detecting objects by means of an optical sensor
Publication Date: 2023.03.15 LEUZE ELECTRONIC GMBH & CO KG
  • EP4148460A1 patent drawingFigure 1
  • EP4148460A1 patent drawingFigure 2~3
  • EP4148460A1 patent drawingFigure 4a~4b

AI summary

The invention relates to an optical sensor (1) for detecting objects (12) with a transmit/receive unit (2), comprising a transmitter (3) emitting transmit light beams (8) and a receiver (4) receiving receive light beams (9), and with an evaluation unit (5) in which a temporal sequence of measured values ​​is generated depending on received signals from the receiver (4), wherein the measured values ​​are distance values ​​or position values. At least one adaptive Kalman filter, in which the measured values ​​are filtered, is provided.