Wyniki wyszukiwania - BazTech

1

Low-cost, low-resolution IR system with super-resolution interpolation of thermal images for industrial applications

Więcek P., Sankowski D.

Measurement Automation Monitoring

|

2018

|

Vol. 64, No. 4

108--111

EN

In this paper authors present application of deep neural networks for super-resolution interpolation of infrared images. A residual neural network with reduced number of layers was used. The transfer learning using RGB visual images was applied in this research. The validation of the network was performed for 32×24 and 160×120 pixels infrared images, with the up-sampling scale factors 2, 3, 4, 5 and 6. Monitoring of high temperature industrial processes like inductive heating and thermal hardening is the main application of proposed methods.

2

Adaptive trained filters with SSIM optimization for video upscaling with soft pixel value truncation

Nakonechnys A., Veres Z.

Czasopismo Techniczne. Automatyka

|

2013

|

R. 110, z. 4-AC

89--101

EN

The paper presents the modified model for video upscaling based on the preservation of structural information from the input sequence in the adaptive trained filters. Additionally, it is proposed to make soft truncation of the result pixel value during the interpolation process. The evaluation of the proposed algorithm has shown good resulting for a variety of test sequences. The best results are obtained for a sequence with average bitrate and movement of the scene. Received results prove the algorithm is valuable for upscaling tasks.

PL

W artykule przedstawiono zmodyfikowany model zwiększania rozdzielczości, oparty na przechowywaniu informacji strukturalnej z wejściowej sekwencji w adaptacyjnych filtrach uczących się. Dodatkowo zaproponowano miękkie progowanie wartości pikseli podczas procesu interpolacji. Weryfikacja proponowanego algorytmu wykazała korzystne rezultaty dla różnych sekwencji testowych. Najlepsze wyniki uzyskano w przypadku sekwencji o średniej szybkości transmisji i przy ruchomym obrazie. Wskazują one na korzyści wynikłe ze stosowania algorytmu w zadaniach zwiększania rozdzielczości obrazu.

3

Interpolatory edge detection

Zeng P. F., Hirata T.

Machine Graphics and Vision

|

2001

|

Vol. 10, No. 2

175-184

EN

Conventional edge operators cannot detect fine edges or edges in low-resolution images in a satisfactory way. In this paper, an interpolatory edge operator is proposed to deal with fine edge detection as a supplement. Fast image expansion is implemented by means of B-spline imterpolation. Edges are determined by seeking pixels with a local maximum of first derivative modulus (modulus maximum). The first derivatives can be calculed quickly by translation and substration along x and y directions if images interpolated by B-splines. Experimental results imply that the proposed interpolatory edge operator can satisfactorily solve the fine edge detection problem in cases when coventional edge operators fail to produce a clear output. Satisfacory results can be obtained when images are interpolated with the expansion rate of three.