New approach of estimating edge detection threshold and application of adaptive detector depending on image complexity

Maksimovic, Vladimir; Petrović, Mile; Savić, Dragan; Jakšić, Branimir; Spalević, Petar

doi:10.1016/j.ijleo.2021.166476

Cited by 11 publications

(7 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, as the authors of [27] note, convolutional neural networks work very slowly with high-resolution images and on devices with weak processors. This is explained by the fact that, in order to obtain an acceptable field susceptibility with convolutional layers, it is necessary to use large kernels (for example, 7 × 7 or 9 × 9), or a large number of layers [28], and this requires large computational resources. To avoid this, most existing systems are limited to image sizes smaller than 41 × 41 (pixels).…”

Section: Theoretical Backgroundmentioning

confidence: 99%

Deep Learning Method for Recognition and Classification of Images from Video Recorders in Difficult Weather Conditions

et al. 2022

View full text Add to dashboard Cite

The sustainable functioning of the transport system requires solving the problems of identifying and classifying road users in order to predict the likelihood of accidents and prevent abnormal or emergency situations. The emergence of unmanned vehicles on urban highways significantly increases the risks of such events. To improve road safety, intelligent transport systems, embedded computer vision systems, video surveillance systems, and photo radar systems are used. The main problem is the recognition and classification of objects and critical events in difficult weather conditions. For example, water drops, snow, dust, and dirt on camera lenses make images less accurate in object identification, license plate recognition, vehicle trajectory detection, etc. Part of the image is overlapped, distorted, or blurred. The article proposes a way to improve the accuracy of object identification by using the Canny operator to exclude the damaged areas of the image from consideration by capturing the clear parts of objects and ignoring the blurry ones. Only those parts of the image where this operator has detected the boundaries of the objects are subjected to further processing. To classify images by the remaining whole parts, we propose using a combined approach that includes the histogram-oriented gradient (HOG) method, a bag-of-visual-words (BoVW), and a back propagation neural network (BPNN). For the binary classification of the images of the damaged objects, this method showed a significant advantage over the classical method of convolutional neural networks (CNNs) (79 and 65% accuracies, respectively). The article also presents the results of a multiclass classification of the recognition objects on the basis of the damaged images, with an accuracy spread of 71 to 86%.

show abstract

Section: Theoretical Backgroundmentioning

confidence: 99%

Deep Learning Method for Recognition and Classification of Images from Video Recorders in Difficult Weather Conditions

et al. 2022

View full text Add to dashboard Cite

show abstract

“…These two stages cannot be separated because it is usual that the image environment and condition consists of many objects and colors, so they must be removed and only use shapes and background colors for this segment/part to be not interfered with any other process to be used as character object recognition. (Maksimovic et al, 2021;Musaddid et al, 2019)…”

Section: Digital Image Processingmentioning

confidence: 99%

Implementation of License Plate Recognition Monitoring System using Neural Network on Solar Powered Microcontroller

Ardiansyah

Bianto²

2022

IVRJ

View full text Add to dashboard Cite

One automatic system for monitoring the presence of vehicles in a parking zone is an indispensable mean of an area such as services building, institutions, and other organizations, which accomodated many vehicles. A tracking record is the most important matter when the vehicle leaves the parking area. Manually, one parking officer will be needed to do the job. However, when using such an automated parking system, then this officers' job can be replaced. The vehicle license plate recording system is designed to use electronic-vision devices as a fundamental device for detecting the presence of vehicle. Vehicle license plates are detected using a digital camera which captured by the camera module on the Raspberry-Pi mini-pc microcontroller device, in addition to the detection of ultrasonic sensors that capture the position of vehicle objects. The process of reading vehicle numbers in an intelligent system of artificial neural networks to extracts each character of the license plate so that each number and letter can be recognized. Meanwhile, the detection of the ultrasonic parking sensor is a complementary confirmation indicating the presence of a vehicle object being monitored. The combination of solar power as the power supply for this automatic system is an important set-up that makes the system's electricity able to run independently. This monitoring system is prepared to help increase vehicle security automatically.

show abstract

“…Like image processing itself, there is great interest among researchers in the edge detection in images where there is noise, so many methods have been used to overcome this problem, and more recently by the method of artificial intelligence and neural networks [6]. This paper categorizes images at three levels of complexity, low (Low Details -LD), medium (Medium Details -MD), and high complexity (High Details -HD), which is determined based on spatial information as in [10].…”

Section: System Modelmentioning

confidence: 99%

Impact of noise to edge detection on images of different complexity

Maksimovic

Milošević

Petrović

et al. 2021

SNG

Self Cite

View full text Add to dashboard Cite

In this paper, an analysis of the edge detection over images of different complexity affected by Salt and Pepper, Gaussian and Speckle noise is performed. An analysis was performed for three noise levels, 0.01, 0.05 and 0.1. Over 100 images from the BSD database were used for analysis and each image has a GroundTruth with which an objective assessment of the detected edges was performed using PR and F measures. Five edge detectors Canny, LoG, Sobel, Prewitt and Roberts operator were used. The results are presented graphically. The obtained results show that noise significantly affects the detection of edges. When it comes to Salt and Pepper noise, Canny detector has achieved the best results for all levels of noise and image complexity. With the Speckle noise type for high and medium number of details in the image, Canny also gave the best results, while for low number of details in the image it is the Prewitt operator. When it comes to Gaussian noise, for all three categories of image complexity the best operator is Prewitt.

show abstract

New approach of estimating edge detection threshold and application of adaptive detector depending on image complexity

Cited by 11 publications

References 20 publications

Deep Learning Method for Recognition and Classification of Images from Video Recorders in Difficult Weather Conditions

Deep Learning Method for Recognition and Classification of Images from Video Recorders in Difficult Weather Conditions

Implementation of License Plate Recognition Monitoring System using Neural Network on Solar Powered Microcontroller

Impact of noise to edge detection on images of different complexity

Contact Info

Product

Resources

About