Object detection under strong background using ASEF-based optical correlator

Optical pattern recognition is heavily challenged when the object is embedded with a strong background and distortions. This study introduces a simple yet powerful technique to identify objects under complex backgrounds using the average of synthetic exact filter (ASEF) with optical correlation. The ASEF is prepared using the training datasets and is applied to the query image to yield the correlation output using the optical correlator. Additionally, the log-polar transform has been incorporated with the ASEF correlation filter to sustain rotational distortion of the query object in the input scene. The combination work together to offer a strong foundation for reliable object recognition in dynamic environments