If you can't find an existing solution it's pretty easy to build a basic one in Java. I just wrote one for a client and it took only a couple hours including researching tools.

Apache POI will read the Excel binary. http://poi.apache.org/

JSONObject will build the JSON

After that it's just a matter of iterating through the rows in the Excel data and building a JSON structure. Here's some pseudo code for the basic usage.

FileInputStream inp = new FileInputStream( file );
Workbook workbook = WorkbookFactory.create( inp );

// Get the first Sheet.
Sheet sheet = workbook.getSheetAt( 0 );

    // Start constructing JSON.
    JSONObject json = new JSONObject();

    // Iterate through the rows.
    JSONArray rows = new JSONArray();
    for ( Iterator<Row> rowsIT = sheet.rowIterator(); rowsIT.hasNext(); )
    {
        Row row = rowsIT.next();
        JSONObject jRow = new JSONObject();

        // Iterate through the cells.
        JSONArray cells = new JSONArray();
        for ( Iterator<Cell> cellsIT = row.cellIterator(); cellsIT.hasNext(); )
        {
            Cell cell = cellsIT.next();
            cells.put( cell.getStringCellValue() );
        }
        jRow.put( "cell", cells );
        rows.put( jRow );
    }

    // Create the JSON.
    json.put( "rows", rows );

// Get the JSON text.
return json.toString();

#!/usr/bin/python
import csv, json
csvreader = csv.reader(open('data.csv', 'rb'), delimiter='\t', quotechar='"')
data = []
for row in csvreader:
    r = []
    for field in row:
        if field == '': field = None
        else: field = unicode(field, 'ISO-8859-1')
        r.append(field)
    data.append(r)
jsonStruct = {
    'header': data[0],
    'data': data[1:]
}
open('data.json', 'wb').write(json.dumps(jsonStruct))