This paper presents a new and eflicient method of encoding uniform image @m and lines. Regions and lines are obtained as the result of image segmentation. split and merge image compression. or as the output of line and polygon drawing algorithms. Lines and contours of uniform regions, are encoded using chain-code. The chaincode is obtained in a way that is &cient with respect to bit-rate and produces lossless contour and line encoding. A lossy method for contour etlcoding is also presented.A set of experiments to compare the performame of traditional chain-code contour etlcoding with the improved contour encoding is presented. The results show a reduction of about 50% in the bit-rate with no reconstruction mor. Efficient lossless chain-encoding of region boundariesSimple modi6cations to t k enoodiog cm lead to a significant saving in bit rate. Pavlidis2 proposes a relative chaincode with variable code per direction. The code used in this paper is a fixed length relative code with priority where a lower number denom higher priority. This coding methcd is designed to inaease code redundancy.A cut of almost 50% in the chaincode can be achieved by modifying the traditional approach of contour tracing into a region boundary tracing method. The problem with cantour following is that pixels 011 region contours are also on the boundaries between regions. Hence, the contour following algarithm etlcodes each boundary twice. The change in approach entails that chains does not descibe closed objects. Therefon?. a few adjustment to the encoding scheme are required. First. one cannot use the condition that the ~d i u a t e s ofthe initial pixel and the coordinates of the terminating pixel of achain are the same, as the means to identify the terminaticm ofachain Rather. achain is ended with a relative prioritized code denoting going back to the previous chain element. Second. this approech has an additional overhead due to the r q u h n e n t to include the coordinates of the initial pixel. Now. the same region may be represented by more than one chain and hence more than om initial pixel. Another modification is that instead of attaching the information about a region to it chainade, the information about all the regions in the image proceeds or succeeds the chaincodes of the regions.The coding scheme bas a built-in LZW Compressa~ In a fork poinl. this algorithm checks the compression table. If possible it chooses a route that is an entry in the table in favor of a mute that re@m updating the table. A lossy option that produce smooth edges is also available Experiments and resultsThe different improvements described in this paper has been applied to four images. One of these images has been obtained from line and polygon drawing. Other images are "natural" iinages that have been passed thmngh segmentation. Our findings an?: 1) Unix compression of the regular chain& achieves between 1.5 times compression to 10 times compression. 2) Boundary Tracing with lossless unnpssion is better than contour tracing with lossless compression...

Objectives/Scope: The performance of ML-based rock properties prediction from seismic with limited and sparse well data is very often inadequate. To address this limitation, we propose a novel automatic well log regularization (ALR) method with specially designed feature augmentation strategy to improve the prediction accuracy. The effectiveness of ALR method is showcased on field data in Malay basin where we successfully predict elastic logs with 30% higher accuracy, while using only 28% less training dataset. Methods, Procedures, Process: The ALR workflow (Figure 1): (1) feature selection and augmentation; (2) training and prediction and (3) prediction optimizations. The workflow starts with predicting any logs type which are available at training but not in blind wells using standard ML workflow for all blind wells (Step 1-2). Then, these intermediately predicted logs at blind well were jointly used as input features together with seismic-derived attributes using a specially designed feature augmentation strategy (Step 3). Finally, Step 1and 2 are then repeated to predict the elastic logs using these augmented input features. Results, Observations, Conclusions: The ALR method was applied on an oil/gas field data in Malay basin to predict elastic logs (AI and SI) at five blind wells from seismic data only and compared to the standard ML workflow. Two wells were used as training (28% of all data). The prediction performance of standard ML workflow (Figure 2a) is poor and can only capture general mean values of the actual AI/SI logs. The results of ALR workflow (Figure 2b) shows 30% better prediction performance compared to the standard ML workflow. In general, the background and high-resolution trend are well captured, and the overall prediction performance is improved using the new proposed prediction method. There are conceivably two explanations for this result: a) the background (low frequency) trend of the well log is properly reconstructed in ALR using only using seismic data. This could mainly lie in the ability of augmented features in better learning the uncertain reflection-reception relationship between seismic data and elastic logs, as well as the spatial/time-varying property of seismic data; (b) The ability to learn meaningful nonlinear feature relationship between input (feature) and output (label) variables with little or no supervision seems to work properly using specially designed feature augmentation. Novel/Additive Information: The ALR method is an ML-based pseudo log generation from seismic data using specially designed feature augmentation strategy. The novel ALR implementation relaxes the requirement of having a massive amount of high-quality labeled data for training and can therefore be applied in areas with limited well data information. ALR method is proven to be highly accurate for direct elastic logs prediction and can potentially be extended to estimate petrophysical properties from seismic data.