Table 1.

Comparisons of the proposed and the previously researched methods.

Category	Methods		Advantage	Disadvantage
Single camera-based	AdaBoost cascade [17]		-Faster processing speeds. -Better performance under low image resolutions.	-Affected by various environmental changes, such as, changing lighting and shadows, and cases where the background temperature is similar to that of the pedestrians’ body.
	HOG–SVM [18,22], integral HOG [19], neural network based on receptive fields [20], and background generation [21]		-More resilient in simple conditions. -Faster processing speed than multiple camera-based algorithm.
	CNN-based method [6,10]		More accurate than the past single camera-based method.
Multiple camera-based	Stereo visible light cameras	Shape and texture information [23]	Better detect pedestrians as it is able to utilize more information than the single camera-based method.	-Longer time to process as it has to process both the camera images.
	Visible light & NIR cameras	HOG-SVM [22]
	Visible light & FIR cameras	Tetra-vision-based HOG-SVM [24]	Better night vision pedestrian detection inside the car.	-No performance without vehicle headlight. -High number of calculation is required as it needs to process two camera images.
		Camera selection [11]	Better performance under various conditions.	-Detection capability is affected as it has no final verification process for the detected pedestrian area.
		Difference image-based fusion [12]
		Late fusion CNN-based method [13]	Higher CNN-based detection accuracy.	-Processing hours and algorithm complexity increases as the method processes input from two camera images to conduct CNN twice.
		Proposed method	-Increased detection reliability (compared to the single camera-based method) by means of adaptively selecting one candidate between two pedestrian candidates received from visible light and FIR camera images. Applies a FIS, and reduces algorithm complexity and processing time. -More resilient detection capability under various environmental changes by means of intensively training and using a diverse dataset.	Design of the fuzzy rule tables and membership function is needed for the FIS.