Body-Based Gender Recognition Using Images from Visible and Thermal Cameras

Area of Science:

• Computer Vision
• Biometrics
• Machine Learning

Background:

• Gender recognition is crucial for various computer vision applications like surveillance and human-computer interaction.
• Existing methods using visible light images are often hindered by environmental factors such as shadows and illumination.
• Robust gender recognition is needed to overcome the limitations of traditional visible-light-based approaches.

Purpose of the Study:

• To develop and evaluate a novel gender recognition method.
• To enhance the accuracy and robustness of gender recognition systems.
• To address the performance degradation issues in visible light-based gender recognition.

Main Methods:

• A new gender recognition method combining visible light and thermal camera images of the human body was proposed.
• Various feature extraction and data fusion techniques were explored and applied.
• Performance was evaluated through comparative analysis against conventional systems.

Main Results:

• The proposed method demonstrated efficiency in gender recognition.
• Experimental results showed improved recognition rates compared to conventional systems.
• The fusion of visible and thermal imaging proved effective for gender recognition.

Conclusions:

• The combined visible and thermal imaging approach offers a more robust solution for gender recognition.
• This method effectively overcomes the limitations posed by varying illumination and shadow conditions.
• The findings suggest significant potential for practical applications in computer vision.