An Analysis and Implementation of the Harris Corner Detector

Sánchez, Javier; Monzón, Nelson; Salgado, Agustín

doi:10.5201/ipol.2018.229

Cited by 65 publications

(24 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In order to perform the translation operation, the interest points, such as corners, are determined using the Harris corner detector operator ( R ) [ 25 , 26 ]. Corners are the key features of the morphology of brain images.…”

Section: Methodsmentioning

confidence: 99%

Combining Sparse and Dense Features to Improve Multi-Modal Registration for Brain DTI Images

Moldovanu

Toporaș

Biswas

et al. 2020

Entropy

View full text Add to dashboard Cite

A new solution to overcome the constraints of multimodality medical intra-subject image registration is proposed, using the mutual information (MI) of image histogram-oriented gradients as a new matching criterion. We present a rigid, multi-modal image registration algorithm based on linear transformation and oriented gradients for the alignment of T2-weighted (T2w) images (as a fixed reference) and diffusion tensor imaging (DTI) (b-values of 500 and 1250 s/mm2) as floating images of three patients to compensate for the motion during the acquisition process. Diffusion MRI is very sensitive to motion, especially when the intensity and duration of the gradient pulses (characterized by the b-value) increases. The proposed method relies on the whole brain surface and addresses the variability of anatomical features into an image stack. The sparse features refer to corners detected using the Harris corner detector operator, while dense features use all image pixels through the image histogram of oriented gradients (HOG) as a measure of the degree of statistical dependence between a pair of registered images. HOG as a dense feature is focused on the structure and extracts the oriented gradient image in the x and y directions. MI is used as an objective function for the optimization process. The entropy functions and joint entropy function are determined using the HOGs data. To determine the best image transformation, the fiducial registration error (FRE) measure is used. We compare the results against the MI-based intensities results computed using a statistical intensity relationship between corresponding pixels in source and target images. Our approach, which is devoted to the whole brain, shows improved registration accuracy, robustness, and computational cost compared with the registration algorithms, which use anatomical features or regions of interest areas with specific neuroanatomy. Despite the supplementary HOG computation task, the computation time is comparable for MI-based intensities and MI-based HOG methods.

show abstract

Section: Methodsmentioning

confidence: 99%

Combining Sparse and Dense Features to Improve Multi-Modal Registration for Brain DTI Images

Moldovanu

Toporaș

Biswas

et al. 2020

Entropy

View full text Add to dashboard Cite

show abstract

“…Didapatkan hasil akurasi diatas 97% dan tingkat presisi objek sebesar 100%. Pengunaan metode Harris Corner akan menghasilkan nilai yang konsiten dari suatu citra walaupun telah mengalami rotasi, penskalaan, variasi pencahayaan, ataupun memiliki noise seperti pada penelitian sebelumnya [4]. Kesimpulannya bahwa metode Hough Transform dan Harris Corner tersebut layak untuk digunakan dalam proses cropping bagian plat nomor kendaraan.…”

Section: Pendahuluanunclassified

“…Metode ini digunakan untuk mendeteksi sudut pada citra, dimana pada sistem ini metode Harris Corner bertujuan untuk mendeteksi sudut pada plat nomor kendaraan yang akan menandakan bahwa plat tersebut berbentuk persegi panjang yang sempurna dengan adanya sudut-sudut tersebut. Metode ini akan menghasilkan nilai yang konsiten dari suatu citra walaupun telah mengalami rotasi, penskalaan, variasi pencahayaan, ataupun memiliki noise [4]. Untuk melakukan proses ini dilakukan langkah-langkah sebagai berikut: 1) Konversi citra biner ke grayscale Hasil keluaran Hough Transform berupa citra biner, maka dari itu harus diubah ke dalam citra grayscale dengan persamaan berikut = 255 * .…”

Section: E Harris Cornerunclassified

Deteksi Plat Nomor Kendaraan dengan Hough Transform dan Harris Corner Menggunakan Akuisisi Melalui Raspberry Pi

Romadhon

2020

TELKA

View full text Add to dashboard Cite

Dalam suatu tempat, seperti di mall, tempat parkir kendaraan dapat terdiri dari beberapa lantai di basement atau di bagian atas bangunan. Di samping itu juga banyak sektor yang berubah fungsi menjadi tempat parkir. Dalam situasi ini, diperlukan sistem monitoring kendaraan yang fleksibel. Pada penelitian ini dibuat sistem yang dapat mengenali jenis kendaraan berdasarkan warna dasar pada plat nomor apakah kendaraan tersebut merupakan kendaraan pribadi, umum, atau pemerintah. Data berupa citra diambil dengan menggunakan webcam melalui akuisisi Raspberry Pi. Sistem dirancang dengan menggunakan deteksi tepi dan morfologi serta menggunakan metode Hough Transform untuk memperbaiki garis tepi dan Harris Corner untuk mendeteksi sudut pada citra. Selanjutnya dilakukan proses cropping plat. Deteksi warna dasar plat nomor dilakukan dengan menggunakan ruang warna YcbC. Selain itu, dilakukan perhitungan waktu proses akuisisi dan waktu proses keseluruhan ketika Raspberry Pi dan laptop melakukan proses deteksi. Berdasarkan hasil pengujian, sistem mendapatkan hasil terbaik sebesar 100% saat deteksi plat nomor kendaaraan pribadi dan hasil terburuk sebesar 70% saat deteksi plat nomor kendaraan umum. Rata-rata akurasi yang didapatkan pada sistem ini sebesar 88,9%. Akurasi tersebut diperoleh dengan 90 data uji. Rata-rata waktu komputasi adalah 0,99 detik, yang artinya proses pendeteksian dilakukan cukup cepat. In some place, such as in a mall, a vehicle parking lot can consist of several floors in the basement or at the top of a building. In addition, many sectors have also been turned into parking lots. In this situation, a flexible vehicle monitoring system is needed to detect the type of vehicle ownership through the base color on the license plate connected to the Raspberry Pi, which is expected to be able to mitigate this problem. This research develop a system that can recognize the type of vehicle based on the color of the plate number whether the vehicle is a private, public, or government vehicle. The data consists of images taken using a webcam through the acquisition of Raspberry Pi. The system was designed by using edge detection, morphology, the Hough Transform method to correct the edge, and Harris Corner to detect angles in the image, after which the plate cropping process is carried out. The license plate base color is detected using the YCbCr color space, and the system will also measure the quality of the network between the Raspberry Pi and the laptop used during the detection process. The system gets the best results at 100% when detecting private vehicle license plates and the worst results at 70% when detecting public vehicle license plates. The average accuracy obtained in this system is 88,9% that used 90 test data. The average computation time is 0,99 seconds, which means that the detection process is quite fast.

show abstract

“…We aimed to resolve the problems of result variations caused by user bias, automation difficulties, and the increased uncertainty incurred when hyperspectral imaging is used to apply the image registration technique. To maintain result consistency and reduce uncertainty, we applied image displacement tracking technology, based on the optical flow algorithm, and Harris corner detection technology [ 26 , 27 ] for extracting corners. It has been reported that optical flow can reduce the uncertainty caused by user inputs and is less affected by rotation and scale than template matching.…”

Section: Introductionmentioning

confidence: 99%

Development of an Image Registration Technique for Fluvial Hyperspectral Imagery Using an Optical Flow Algorithm

You

Kim

2021

Sensors

View full text Add to dashboard Cite

Fluvial remote sensing has been used to monitor diverse riverine properties through processes such as river bathymetry and visual detection of suspended sediment, algal blooms, and bed materials more efficiently than laborious and expensive in-situ measurements. Red–green–blue (RGB) optical sensors have been widely used in traditional fluvial remote sensing. However, owing to their three confined bands, they rely on visual inspection for qualitative assessments and are limited to performing quantitative and accurate monitoring. Recent advances in hyperspectral imaging in the fluvial domain have enabled hyperspectral images to be geared with more than 150 spectral bands. Thus, various riverine properties can be quantitatively characterized using sensors in low-altitude unmanned aerial vehicles (UAVs) with a high spatial resolution. Many efforts are ongoing to take full advantage of hyperspectral band information in fluvial research. Although geo-referenced hyperspectral images can be acquired for satellites and manned airplanes, few attempts have been made using UAVs. This is mainly because the synthesis of line-scanned images on top of image registration using UAVs is more difficult owing to the highly sensitive and heavy image driven by dense spatial resolution. Therefore, in this study, we propose a practical technique for achieving high spatial accuracy in UAV-based fluvial hyperspectral imaging through efficient image registration using an optical flow algorithm. Template matching algorithms are the most common image registration technique in RGB-based remote sensing; however, they require many calculations and can be error-prone depending on the user, as decisions regarding various parameters are required. Furthermore, the spatial accuracy of this technique needs to be verified, as it has not been widely applied to hyperspectral imagery. The proposed technique resulted in an average reduction of spatial errors by 91.9%, compared to the case where the image registration technique was not applied, and by 78.7% compared to template matching.

show abstract

An Analysis and Implementation of the Harris Corner Detector

Cited by 65 publications

References 19 publications

Combining Sparse and Dense Features to Improve Multi-Modal Registration for Brain DTI Images

Combining Sparse and Dense Features to Improve Multi-Modal Registration for Brain DTI Images

Deteksi Plat Nomor Kendaraan dengan Hough Transform dan Harris Corner Menggunakan Akuisisi Melalui Raspberry Pi

Development of an Image Registration Technique for Fluvial Hyperspectral Imagery Using an Optical Flow Algorithm

Contact Info

Product

Resources

About