Automatic segmentation and recognition of unconstrained handwritten numeral strings
Bibliographic record
Abstract
Segmentation and recognition of handwritten numeral strings is a very interesting and challenging problem in pattern recognition. It also has a lot of important applications such as: postal code recognition, bank check processing; tax form reading, etc. In this thesis, a new system for the segmentation and recognition of unconstrained handwritten numeral strings is proposed. The system uses a combination of foreground and background features for the segmentation of touching numerals in strings. The method introduces new algorithms for the traversal of top and bottom foreground and background skeletons, and top and bottom contours of numerals. Then; it tries to locate all feature points on these skeletons and contours and alternatively match feature points from top to bottom (or bottom to top) of the images to build all possible candidate segmentation paths (so-called segmentation hypotheses). A novel genetic representation scheme is utilized in order to represent the space of all possible segmentation hypotheses. In order to improve searching and evolution of segmentation hypotheses and facilitate finding the ones with the highest confidence values of segmentation and recognition, this genetic framework utilizes contextual knowledge extracted from string images. A novel evaluation scheme based on segmentation and recognition scores is introduced in order to improve the evaluation of segmentation hypotheses and to enhance the outlier resistance of the system. In order to improve stability and plasticity of our system in the learning and recognition of numerals, a new algorithm for clustering of handwritten digits based on their shapes is proposed. Also, in order to improve the searching power of our system and its convergence, a new evolutionary algorithm based on genetic particle swarm optimization (GBPSO) is proposed. Numerous experiments using images from well known databases of handwritten numeral strings such as CENPARMI, NIST NSTRING SD19, and our newly created databases of Farsi/Arabic numerals have been conducted in order to evaluate the performance of the proposed method. Experiments have shown that proper use of contextual knowledge in segmentation; evaluation and search greatly improves the overall performance of the system. This system shows superior results compared with those reported in the literature.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".