The Cheque Detection dataset is sourced from the IDRBT Cheque Image Dataset and features a diverse range of textures and ink colors. It comprises a total of 112 cheque leaves from four different Indian banks. The dataset is organized into seven classes, including sign, dateiss, chqno, amt, acno, receivername and issuebank which represent various elements commonly found on cheques. This dataset is particularly valuable for tasks related to automated cheque processing, such as signature verification, date extraction, and the extraction of relevant information such as check numbers, amounts, account numbers, payee names, and issuing banks.

In order to simulate the pen ink difference in cheque leaves, seven blue and seven black pens are used. To avoid bias due to writing, nine different volunteers have taken active participation in preparing the data set. A total of 14×9 = 126 pen−volunteer combinations (fourteen pens and nine volunteers) are used for pen ink data generation. In a practical scenario, similar color pens are used for the addition of new words in the source document. Each cheque is written by two volunteers using two different pens (either blue or black). Hence, the data set is created with 42 possible combinations of blue and black pens. All the cheque leaves are scanned in the normal scanner at 300 dpi resolution. Handwritten words from each scanned cheque are cropped manually and grouped based on the pen used to write the words.

The authors have introduced a novel method for identifying pen ink differences in handwritten bank checks. They frame the problem as a binary classification task and extract thirty features for each word pair instance. These features serve as the basis for training a Multi-Layer Perceptron (MLP) classifier using known pen ink data. The performance of the MLP classifier is assessed on both known and unknown pen ink pixels. A thorough analysis and comparison of the results demonstrate the superior performance of the proposed method when compared to existing approaches, particularly for distinguishing between black and blue pen samples in handwritten checks.

Forensics in handwritten Bank cheques from the perspective of differentiating pen ink has great importance to the judicial system. In handwritten Bank cheque forensics, it is often important to establish a relation between the pen inks. It helps to identify whether a single pen has been used to write the Bank cheque or multiple pens. Numerous possibilities of fraud exist in handwritten Bank cheques. In this work, authors focus only on pen ink differentiation in Bank cheques. Possibilities of fraud in any Bank cheque and its consequences help to understand the importance of the work.

An example of new word addition in a Bank cheque using a different pen is depicted in the Figure below which is elaborated as follows. The cheque was initially issued to Mr. Ravi Kumar Singh, amounting to Seventy thousand only. Later, the forger appended new words in the pay name and amount section as marked by red circles. This difference in pen ink can not be always perceived by the naked eye. This type of case helps to understand the possibility of the addition of new words in handwritten Bank cheques. A number of handwritten document frauds are possible in bills, business agreements, educational documents, etc. This motivates authors to differentiate pen ink in Bank cheques.

Addition of handwritten words (marked by red circles) in a Bank cheque image.