Dr. Chen Yang Profile

Dr. Chen Yang

at Science and Technology on Optical Radiation Lab

SPIE Involvement:

Author

Publications (3)

Proceedings Article | 12 December 2018 Paper

Condition detection of overhaul platform based on image processing

Yajun Song, Chen Yang, Jinbao Yang, Yachao Liu

Proceedings Volume 10846, 108461F (2018) https://doi.org/10.1117/12.2504741

KEYWORDS: Cameras, Image processing, Intelligence systems, Manufacturing, Image classification, Video surveillance, Detection and tracking algorithms, Imaging systems, Feature extraction, Manufacturing equipment

Read Abstract +

EMU(Electric Multiple Units) maintenance platform is one of the key equipments of the motor Train base and EMU application, provides convenient for EMU daily maintenance and inspection, maintenance and repair. In the process design of the equipment, hydraulic crossover and other methods are adopted to meet the needs of the maintenance personnel on different types of EMUs. Therefore, the tipping boards are set up on the actual platform and protection network to ensure the safety of the workers' reaching the top. The EMUs usually use a four line storehouse, and the number of tipping plates is hundreds of them. Most of them are reversing / receiving by cylinder driving, and a large number of flip plates are distributed in the workshop. Besides, the state of the retractable boards and the faults have important influence on the operation. Therefore, it is very important to monitor the state of the flip plate. At present, it is more dependent on the monitoring and controlling system of the mechanical stroke switch in China. Due to the state of the hardware and the stability of the system, the method has the risk of omission and false alarm in some cases. Meanwhile, the switch signal can not directly display the real-time state of the tipping board. Therefore, the present invention provides a high-definition network camera image processing board state detection system based on double plate state real-time monitoring and detection, enhance the safety of EMU maintenance platform. The real-time image acquisition, calculating the specified texture information area gray co-occurrence matrix, generating feature vector descriptor, using support vector machine (SVM) classification method for feature vector classification, training classifier, and use the classifier to recognize the flap, and locate the flap position. And SVM is a novel method of machine learning evolving from statistics. SVM presents many own advantages in solving machine learning problems such as small samples, nonlinearity and high dimension. In this paper, the image texture SVM classification method construct feature vectors through the extraction of image gray level co-occurrence matrix texture information.The metioned texture imformation include gray-level co-occurrence matrix energy, contrast and entropy. And the gray level co-occurrence matrix reflects the image direction, adjacent interval and the change in value of integrated information.The results show that the image texture SVM classification method can effectively combine the method of template matching for the identification of condition detection of overhaul platform.

Proceedings Article | 24 October 2017 Paper

Identification of conveyor belt injury based on image texture SVM classification method

Chen Yang, Yajun Song, Jinbao Yang, Yachao Liu

Proceedings Volume 10462, 104620D (2017) https://doi.org/10.1117/12.2282029

KEYWORDS: Image classification

Read Abstract +

In the field of steel wire core conveyor belt detection, the X-ray detection method has been widely used for its accuracy and reliability. But as a result of the monotonous, baldness and plenty of X-ray image, it is necessary to apply the computer image processing techniques to identification wire injury accurately and automatically. Then the injury can be located in the actual conveyor belt to maintenance and repair. And the joint is an important part of the conveyor belt, for it is easy to recognize in the practical application, so as the benchmark to joint reference point positioning fault is a good choice. Therefore, the accurate identification of joint is very important to conveyor belt injury locational. At present there are some algorithms applying the method of detecting domain gray level or horizontal gradient change frequency in the identification of joint. These algorithms can be accurate to single image of joint without considering the practical complex and changeable X-ray image for the different thickness of the conveyor belt outer rubber. For single image characteristics of the proposed algorithm is easy to failure in practical application. A new robust algorithm is necessary to solve this problem. And SVM(Support Vector Machine) is a novel method of machine learning evolving from Statistics. SVM presents many own advantages in solving machine learning problems such as small samples, nonlinearity and high dimension. In this paper, the image texture SVM classification method construct feature vectors through the extraction of image gray level co-occurrence matrix texture information. classified feature vectors using the SVM classification method to determine whether the image contains joint and provide the joint location information. The mentioned texture information include gray-level co-occurrence matrix energy, contrast and entropy. And the gray level co-occurrence matrix reflects the image direction, adjacent interval and the change in value of integrated information. SVM classification method is applied to locate the joints number and position on the real conveyor belt. And by means of image binarization, skeleton and such as pretreatments, this paper use the method of template matching for the identification wire fracture. Finally the method locate injury on the real conveyor belt according to the fracture position and joint position of the pixel distance. The results show that the image texture SVM classification method can effectively combine the method of template matching for the identification of conveyor belt injury.

Proceedings Article | 24 October 2017 Paper

Visual and infrared image fusion algorithm based on adaptive PCNN

Yajun Song, Chen Yang, Zhi Chai, Jinbao Yang

Proceedings Volume 10462, 1046239 (2017) https://doi.org/10.1117/12.2285118

KEYWORDS: Infrared imaging, Image fusion, Neural networks

Read Abstract +

As the third generation artificial neural network, pulse coupled neural network (PCNN) which consider the characteristics of neurobiology of time coding and spatial accumulation, getting incomparable advantages comparing with the traditional artificial neural network, has broad application prospects in image fusion. In recent years, improving traditional model and adaptive adjustment of key parameters of the model have become major focuses gradually. In this paper, a novel visual and infrared image fusion algorithm is presented based on a new modified PCNN model. The key parameter of linking strength of the model is calculated with the character of the input images adaptively. Firstly, the modified PCNN employs index map and threshold look-up table to improve traditional PCNN model. Threshold look-up table records the thresholds which correspond to the different iteration layers of the modified PCNN model. To improve the computing speed of modified model, the thresholds could be calculated before the modified model starts to compute, which reduces the computing burden of traditional model to get the thresholds. Index map records the firing time of the input image’s pixels during modified PCNN model computing. The values of index map represent the integrating results of similar pixels in space neighborhood of the input image, which reflect the global visual features of the input image. Then, aiding method is used to compute the value of linking strength of modified PCNN model. The linking strength represents the degree that the linking input modulates the feeding input of the current neuron. If the value of linking strength can be decided in accordance with the specific characteristics of the input images, better fusion performance should be gotten in theory. Considering visual image has more detail information of target and infrared image has more energy character of target, local entropy and local energy are combined with the linking strength parameter of modified PCNN model for visual and infrared image separately in the proposed method of this paper. Finally, original visual and infrared image are processed with the modified PCNN model by calculating the linking strength using above procedure. The image fusion rules based on the index maps of visual and infrared image are used to calculate the fusion image. In order to evaluate the performance of the proposed method, a large number of experiments are made. In the experiments, the typical image sets which selected in many related papers are processed with the proposed method and wavelet transform separately. The different fusion images are evaluated with subjective and objective criteria, including the average, standard deviation and spatial frequency. Average stands for average value of pixel’s gray level. Standard deviation manifests that discrete situation for gray level related to average value. Spatial frequency could measure the image details information. The calculated results shows that, compared to methods like wavelet transform, the proposed method can improve the objective criteria values significantly.

View contact details

UPDATE YOUR PROFILE

Is this your profile? Update it now.

Sign into your SPIE.org account

Don’t have a profile and want one?

Create an account on SPIE.org

Keywords/Phrases

Search In:

Publication Years