Vol.:(0123456789) 1 3 Evolutionary Intelligence https://doi.org/10.1007/s12065-018-0160-6 SPECIAL ISSUE Crow search algorithm with discrete wavelet transform to aid Mumford Shah inpainting model Balasaheb H. Patil 1 · P. M. Patil 2 Received: 25 April 2018 / Revised: 6 July 2018 / Accepted: 12 July 2018 © Springer-Verlag GmbH Germany, part of Springer Nature 2018 Abstract Inpainting plays a signiﬁcant role in solving a variety of image processing issues that comprises zooming, removal of impulse noise, removal of scratches etc. These speciﬁed signiﬁcances are all associated to inpainting in image domain. Even though more advanced inpainting models have been introduced, it suﬀers from problem of having low quality. Hence this paper intends to develop a novel inpainting model on the basis of MS modeling. Initially, the pre-processing of the image is done by Discrete Waveley Transform (DWT) and further, its given to MS inpainting model. Moreover, the ﬁlter coeﬃcient in DWT algorithm is optimized by Crow Search Algorithm (CSA), that is being considered as the main objective. As the result- ant image involves more scratches, this proposed model necessitates smoothening image model using Reproducing Kernel Hilbert Smoothing (RKHS). With all these techniques, the proposed inpainting model is termed as Crow Search Optimized DWT Kernel-based MS (CODWTK-MS). During the performance analysis, the proposed method is compared over various traditional inpainting models like MS, DWT-based MS, DWT Kernel-based MS, and Dragonﬂy Optimized DWT Kernel- based MS (DODWTK-MS) in terms of several measures and proves the superiority of proposed inpainting model. Keywords Image inpainting · Mumford Shah model · Discrete wavelet transform · Filter coeﬃcient · Crow search optimization 1 Introduction Inpainting addresses the issues in ﬁlling the missing image parts, and it is designed owing to certain factors like, from eliminating scrapes in photos, reinstating prehistoric sketches, and ﬁlling in the lost pixels of images conveyed by means of noisy channels [1–3].In the imaging ﬁeld, the name “inpainting” refers to the recuperating of images with loosed or missing data [4]. It remains as a signiﬁcant work in a variety of image reinstatement issues together with impulse noise removal, zooming, and scratch removal, and so on. In the image domain, these stated signiﬁcances are all associ- ated with the inpainting technique [5]. On the other hand, in the actual transmission and storage of the digital pixels, the DWT is an extremely admired technique for the coding of sparse [6]. The region with lost parts is said to be the area of inpainting [7], and it must be ﬁlled in such a manner that both texture information and edges (structure) [8] of the image stay consistent. Digital inpainting could be adopted in numerous applica- tions, like, removal of objects, image restoration, or to raise the resolution of the image. The research is concentrating on ﬁnishing areas with missing details which have been miss- ing or removed deliberately. Performers counterfeited the conception of inpainting technique by means of their indi- vidual awareness and capabilities for renovating or ﬁxing reimbursements in monuments or paintings. Nowadays, the capability of numerous digitalizing kinds of visual infor- mation generates the requirement for methods that further refurbish digital malfunctions, as performed with paintings [9]. On considering several video, communication or audio applications, it has been required to recuperate a signal ruined by narrowband intervention, like electric hum [10]. These kinds of interventions can be obviously meagrely symbolized in the domain of frequency [11]. * Balasaheb H. Patil balasahebhpatil144@gmail.com P. M. Patil b_ash11@rediﬀmail.com 1 Sinhgad College of Engineering, Pune, Maharashtra, India 2 Jayawantrao Sawant College of Engineering, Pune, Maharashtra, India