Traditional Retinex model-based image enhancement methods require careful design of constraints and parameters to handle this highly ill-conditioned decomposition. With the advancement of deep learning algorithms, low-light image enhancement has also achieved deep processing. However, image enhancement based on the RGB color space model is prone to color distortions when enhancing images under the influence of the correlation of the three primary colors of RGB. In this report we apply the HSV color space technique to a Retinex-based network model. Simulations and experiments show that using HSV space-improved deep neural networks can effectively avoid the color distortion problem.
The performance of the imaging system under low light intensity will be affected by shot noise, and the shot noise will become stronger as the power of the light source decreases. Aiming at the impact of shot noise, this paper applies the principle of deep learning to low-light image enhancement. To improve the generalization ability of deep neural networks in different scenarios, a block matching solution based on BM3D is proposed to optimize the data of the Retinex network model. In the training process of the network, the consistency of the reflection component and the smoothness of the illumination component of the low-light image and the normallight image are used to constrain, without the need for real data of the reflection component and the illumination component. Experimental results show that this method can obtain a satisfactory low-light enhancement effect, and can significantly improve the reconstruction results of low-light images affected by noise.
A fast and precise spatial-carrier phase-shifting algorithm based on the matrix VU factorization strategy that can realize dynamic real-time phase detection is proposed. First, the proposed algorithm divides the spatial-carrier interferogram into four phase-shifting subinterferograms. Second, the matrix VU factorization strategy, an excellent fast iterative algorithm, is used to accurately obtain the measured phase from these subinterferograms. Numerical simulation and experimental comparison verify that this method is an efficient and accurate single-frame phase demodulation algorithm. Meanwhile, the performance of the proposed method is analyzed and discussed for the influencing factors, such as random noise level, carrier-frequency value, and carrier-frequency direction. The results show that the method proposed is a fast and precise phase detection method that provides another effective solution for dynamic real-time phase measurement.
Transport of Intensity Equation (TIE) is a simple and efficient method for phase retrieval by solving the equation between the intensity axial derivative and phase. In this method, the estimation of the axial derivative of intensity is very crucial. Simply, we use two defocused intensity images to estimate the axial derivative by finite difference method. However, the result is still unsatisfactory even though the optimal defocused distance is adopted. The reason lies in that the intensity’s axial change is not linear in the propagation of light. Simply using the finite difference between the two defocused images will ignore higher order axial derivatives. In other words, the estimation of the axial derivative of intensity will contain nonlinear errors. To solve this problem, we propose an extrapolation-based method to estimate the axial derivative of intensity using multiple intensity images. With Taylor expansion and a series of combination and eliminations on these images, high order terms of axial derivative errors are removed. As a result, the nonlinear errors in estimation of the axial derivative will be reduced. The performance of our proposed method for different types of phases under different illumination conditions is investigated. Compared with normal TIE, our method can obtain a much more accurate phase profile.
The extreme value of interference (EVI) algorithm is a very fast and efficient method for the fringe pattern phase demodulation. It requires only two arbitrarily phase-shifted frames in which the phase shift between interferograms can be determined by searching the maximum and the minimum of the normalized interference patterns, then the measured phase is obtained by an arctangent function. Compared with other two-frame demodulation algorithms, the EVI algorithm has great advantages. Firstly, the EVI algorithm is simple and the calculation speed is fast. Secondly and more importantly, it works very well even if the number of fringes of the interferogram is less than one. However, to make this method work, the fringe should be normalized in advance, which is sometimes not a satisfactory requirement. The effects of uneven background terms, modulation amplitude variations, and random noise in the fringe pattern will make the normalization of the fringes extremely complex. Therefore, by employing the HilbertHuang transform (HHT) based prefiltering in this paper, the background intensities and modulation amplitudes of the two interferograms are suppressed and normalized respectively. Then, phase demodulation is implemented using the EVI method. Because of the HHT process, the demodulation result is greatly improved in plenty of situations. Both simulation and experimental studies have shown that the proposed improved method makes it easier to determine the phase distribution with high precision even under complex conditions.
Annular sub-aperture stitching interferometry (ASSI) has provided an alternative solution to measure rotationally symmetric aspheric surfaces with low cost and high flexibility. It is an effective way to test the aspheric surface with a larger aperture and larger relative aperture without null compensation. In this paper, two kinds of annular sub-aperture stitching algorithms, pairwise sequential stitching (PSS) and global synchronously stitching (GSS), were studied. The detailed mathematical expressions are shown in the form of matrix. Besides, the influence of the noise and number of sub-apertures to the two algorithms was also studied by simulation. At last, experimental results of a convex hyperboloid surface by using the two stitching algorithms are presented.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.