Influence of Image Enhancement Techniques on Effectiveness of Unconstrained Face Detection and Identification
Само за регистроване кориснике
2021
Чланак у часопису (Објављена верзија)
Метаподаци
Приказ свих података о документуАпстракт
In a criminal investigation, along with processing forensic evidence, different investigative techniques are used to identify the perpetrator of the crime. It includes collecting and analyzing unconstrained face images, mostly with low resolution and various qualities, making identification difficult. Since police organizations have limited resources, in this paper, we propose a novel method that utilizes off-the-shelf solutions (Dlib library Histogram of Oriented Gradients-HOG face detectors and the ResNet faces feature vector extractor) to provide practical assistance in unconstrained face identification. Our experiment aimed to establish which one (if any) of the basic image enhancement techniques should be applied to increase the effectiveness. Results obtained from three publicly available databases and one created for this research (simulating police investigators’ database) showed that resizing the image (especially with a resolution lower than 150 pixels) should always precede ...enhancement to improve face detection accuracy. The best results in determining whether they are the same or different persons in images were obtained by applying sharpening with a high-pass filter, whereas normalization gives the highest classification scores when a single weight value is applied to data from all four databases.
Кључне речи:
Face detection / Face recognition / Histogram of oriented gradients / Image processingИзвор:
Elektronika Ir Elektrotechnika, 2021, 27, 5, 49-58Издавач:
- Kaunas : Kauno Technologijos Universitetas
Институција/група
JakovTY - JOUR AU - Vuković, Igor AU - Čisar, Petar AU - Kuk, Kristijan AU - Banđur, Miloš AU - Popović, Brankica PY - 2021 UR - http://jakov.kpu.edu.rs/handle/123456789/1522 AB - In a criminal investigation, along with processing forensic evidence, different investigative techniques are used to identify the perpetrator of the crime. It includes collecting and analyzing unconstrained face images, mostly with low resolution and various qualities, making identification difficult. Since police organizations have limited resources, in this paper, we propose a novel method that utilizes off-the-shelf solutions (Dlib library Histogram of Oriented Gradients-HOG face detectors and the ResNet faces feature vector extractor) to provide practical assistance in unconstrained face identification. Our experiment aimed to establish which one (if any) of the basic image enhancement techniques should be applied to increase the effectiveness. Results obtained from three publicly available databases and one created for this research (simulating police investigators’ database) showed that resizing the image (especially with a resolution lower than 150 pixels) should always precede enhancement to improve face detection accuracy. The best results in determining whether they are the same or different persons in images were obtained by applying sharpening with a high-pass filter, whereas normalization gives the highest classification scores when a single weight value is applied to data from all four databases. PB - Kaunas : Kauno Technologijos Universitetas T2 - Elektronika Ir Elektrotechnika T1 - Influence of Image Enhancement Techniques on Effectiveness of Unconstrained Face Detection and Identification VL - 27 IS - 5 SP - 49 EP - 58 DO - 10.5755/j02.eie.29081 ER -
@article{ author = "Vuković, Igor and Čisar, Petar and Kuk, Kristijan and Banđur, Miloš and Popović, Brankica", year = "2021", abstract = "In a criminal investigation, along with processing forensic evidence, different investigative techniques are used to identify the perpetrator of the crime. It includes collecting and analyzing unconstrained face images, mostly with low resolution and various qualities, making identification difficult. Since police organizations have limited resources, in this paper, we propose a novel method that utilizes off-the-shelf solutions (Dlib library Histogram of Oriented Gradients-HOG face detectors and the ResNet faces feature vector extractor) to provide practical assistance in unconstrained face identification. Our experiment aimed to establish which one (if any) of the basic image enhancement techniques should be applied to increase the effectiveness. Results obtained from three publicly available databases and one created for this research (simulating police investigators’ database) showed that resizing the image (especially with a resolution lower than 150 pixels) should always precede enhancement to improve face detection accuracy. The best results in determining whether they are the same or different persons in images were obtained by applying sharpening with a high-pass filter, whereas normalization gives the highest classification scores when a single weight value is applied to data from all four databases.", publisher = "Kaunas : Kauno Technologijos Universitetas", journal = "Elektronika Ir Elektrotechnika", title = "Influence of Image Enhancement Techniques on Effectiveness of Unconstrained Face Detection and Identification", volume = "27", number = "5", pages = "49-58", doi = "10.5755/j02.eie.29081" }
Vuković, I., Čisar, P., Kuk, K., Banđur, M.,& Popović, B.. (2021). Influence of Image Enhancement Techniques on Effectiveness of Unconstrained Face Detection and Identification. in Elektronika Ir Elektrotechnika Kaunas : Kauno Technologijos Universitetas., 27(5), 49-58. https://doi.org/10.5755/j02.eie.29081
Vuković I, Čisar P, Kuk K, Banđur M, Popović B. Influence of Image Enhancement Techniques on Effectiveness of Unconstrained Face Detection and Identification. in Elektronika Ir Elektrotechnika. 2021;27(5):49-58. doi:10.5755/j02.eie.29081 .
Vuković, Igor, Čisar, Petar, Kuk, Kristijan, Banđur, Miloš, Popović, Brankica, "Influence of Image Enhancement Techniques on Effectiveness of Unconstrained Face Detection and Identification" in Elektronika Ir Elektrotechnika, 27, no. 5 (2021):49-58, https://doi.org/10.5755/j02.eie.29081 . .